93
ARC Centre of Excellence for Complex Dynamic Systems and Control 2008 ANNUAL REPORT

2008 - University of Newcastle

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: 2008 - University of Newcastle

UoN 2008/1520

ARC Centre of Excellence for Complex Dynamic Systems and Control

2008ANNUAL REPORT

The University of Newcastle Callaghan NSW 2308 Australia

T +61 2 4921 7072 F +61 2 4960 1712 W www.newcastle.edu.au/centre/cdsc

Page 2: 2008 - University of Newcastle

2008 ANNUAL REPORT 01

TABLE Of CONTENTS

OUR VISION 4

DIRECTOR’S REPORT 5

STAFF 6

STAFF MOVEMENTS 6

POSTGRADUATE RESEARCH STUDENTS 7

THESES SUBMITTED IN 2008 8 GRADUATED 2008 8

ADVISORY BOARD 9

VISITORS 10 ACADEMIC 10 STUDENT 11

INDUSTRY INTERACTION AND SELECTED OUTCOMES 2008 12

CONFERENCES, COURSES AND WORKSHOPS 2008 14

SEMINARS 15

SELECTED HIGHLIGHTS 2008 17

RESEARCH PROGRAMS 18

A. INDUSTRIAL CONTROL AND OPTIMISATION 19

A.1 PERfORMANCE OPTIMISATION Of MARINE SYSTEMS 19 A.1.1 Marine system simulation tools 20 A.1.2 Identification of radiation force parametric models of marine structures from 2D frequency domain data 20 A.1.3 Modelling and control of parametric resonance in marine vessels 21 A.1.4 Experiment design and identification of marine vessels dynamics 21 A.1.5 System identification for rapid model prototyping of ship training 22 simulators (Offshore Simulator Centre AS, Norway.) A.1.6 Adaptive control of gyroscopes for roll stabilisation of marine 23 vessels (Halcyon International, Australia) A.1.7 Control design for optimum power extraction in wave energy 24 converters (Robotiker-Tecnalia, Spain.) A.1.8 Constrained predictive control of ship fin stabilizers to prevent 24 dynamic stall

A.2 OPTIMISATION BASED OPERATOR GUIDANCE SCHEMES 25 A.2.1 Modelling and control of copper heap bioleaching processes 25 A.2.2 Sferics reduction in electromagnetic mineral exploration 27 A.2.3 Co-generation at westVAMP 27

A.3 INTEGRATED MINE PLANNING (BHP BILLITON) 28

A.4 NEXT GENERATION MODEL-BASED CONTROL TOOLS (MATRIKON) 30 A.4.1 Next generation model based control tools for CPO 30 A.4.2 Next generation model based control tools for processmore 31

A.5 CSR SUGAR (INDUSTRIAL AffILIATE) 32 A.5.1 Constrained, multi-variable control of an integrated sugar mill system for economic enhancement 32 A.5.2 CSR brake van control 33

A.6 CONNELL WAGNER (INDUSTRIAL AffILIATE) 34

A.7 HATCH (INDUSTRIAL AUTOMATION SERVICES) (INDUSTRIAL AffILIATE) 34

A.8 BOEING RESEARCH AND TECHNOLOGY, AUSTRALIA (INDUSTRIAL AffILIATE) 34

B. MECHATRONICS 35

B.1 CHARGE DRIVES fOR SCANNING PROBE MICROSCOPE POSITIONING STAGES 35

B.2 SIMULATION Of DYNAMICS-COUPLING IN PIEZOELECTRIC TUBE SCANNERS BY REDUCED ORDER fINITE ELEMENT MODELS 36

B.3 MINIMIZING SCANNING ERRORS IN PIEZOELECTRIC STACK-ACTUATED NANOPOSITIONING PLATfORMS 37

B.4 HIGH-BANDWIDTH CONTROL Of A PIEZOELECTRIC NANOPOSITIONING STAGE IN THE PRESENCE Of PLANT UNCERTAINTIES 37

B.5 DESIGN, ANALYSIS AND CONTROL Of A fAST NANOPOSITIONING STAGE 38

B.6 SIMULTANEOUS SENSING AND ACTUATION WITH A PIEZOELECTRIC TUBE SCANNER 39

B.7 IDENTIfICATION AND CONTROL Of NEGATIVE IMAGINARY SYSTEMS 40

B.8 INTEGRAL RESONANT CONTROL Of PIEZOELECTRIC TUBE NANOPOSITIONERS 41

B.9 PRECISE TIP POSITIONING Of fLEXIBLE MANIPULATORS 42

Page 3: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control02

C. CONTROL SYSTEM DESIGN 43

C.1 NONLINEAR MODEL PREDICTIVE CONTROL 43 C.1.1 Predictive power control of wireless sensor networks 43 for closed loop control C.1.2 A vector quantisation approach to scenario generation for 43 stochastic nmpc C.1.3 Constrained nonlinear model predictive control 44

C.2 ROBUST MODEL PREDICTIVE CONTROL 44 C.2.1 Robust output-feedback model predictive control for systems with unstructured uncertainty 44

C.3 VIRTUAL LABORATORIES fOR CONTROL EDUCATION 45

C.4 ACTUATOR fAULT TOLERANT CONTROL 46

C.5 LIMITATIONS IN fEEDBACK CONTROL OVER 47 COMMUNICATION CHANNELS C.5.1 Infimal feedback capacity for a class of additive 47 coloured gaussian noise channels C.5.2 Repeated poles in feedback over a class of signal-to-noise 47 ratio constrained channels C.5.3 Minimum variance control over a gaussian communication channel 48 C.5.4 Disturbance rejection in channel signal-to-noise ratio 48 constrained feeback control C.5.5 Performance limitations in distributed control systems 49

C.6 CONTROL Of POWER PLANTS 50 C.6.1 Performance limitations arising in the control of power plants 50

C.7 ULTIMATE BOUNDS IN PERTURBED SYSTEMS 50

C.8 QUANTISED CONTROL 50

D. SIGNAL PROCESSING 51

D.1 SYSTEM IDENTIfICATION 51 D.1.1 Experiment design for system identification 51 D.1.2 On parameter estimation of the schwartz-smith 51 short-term/long-term model D.1.3 Identifiability of errors in variables dynamic systems 52 D.1.4 Robust identification of process models from plant data 52 D.1.5 Virtual closed loop identification: a generalized tool for 52 identification in closed loop D.1.6 Relative error issues in sampled data models 52 D.1.7 Redundancy vs multiple starting points i nonlinear 52 systems related inverse problems D.1.8 On useful redundancy in robust optimal experiment 52 design for nonlinear system identification

D.2 DUAL-STAGE SYSTEMS 53 D.2.1 Development of a linear motion dual-stage actuator 53 D.2.2 Development of a rotational motion dual-stage actuator 53 D.2.3 Nonlinear tracking control for a hard disk drive 53 D.2.4 Development of an extended reset controller and its 54 experimental demonstration D.2.5 A reset state estimator using an accelerometer for 54 enhanced motion control with sensor quantisation

D.3 QUANTISATION 54 D.3.1 feedback quantisers 54 D.3.2 Quantisation of filter bank frame expansions through 54 moving horizon optimisation

D.4 TIME-fREQUENCY ANALYSIS 55 D.4.1 Linear system modelling in the subband domain 55 D.4.2 High-speed analog-to-digital converter design 56 D.4.3 Efficient sound synthesis for real-time applications 56 D.4.4 Proper group actions in abstract harmonic analysis 56

D.5 COMMUNICATION SYSTEMS 57 D.5.1 A complex-baseband active-set approach for tone reservation 57 par reduction in ofdm systems D.5.2 A multi-step detector for linear isi-channels incorporating degrees 57 of belief in past estimates D.5.3 Em-based receiver design for uplink mimo-ofdma systems 57 D.5.4 A new method for lowering the error floors of non-binary LDPC codes 57 D.5.5 Outer code design for serially concatenated continuous phase 57 modulation with symbol-wise interleading

Page 4: 2008 - University of Newcastle

2008 ANNUAL REPORT 03

E. STATISTICS 58

E.1 PARAMETRIC BAYESIAN MODELLING 59 E.1.1 Spatio-temporal mixture models 59 E.1.2 Image classification 59 E.1.3 Meta-analysis 59

E.2 BAYESIAN NONPARAMETRICS 59

E.3 COMPLEX MODELS 59 E.3.1 Bayesian networks 59 E.3.2 Genetics 60 E.3.3 Environment and health 60

E.4 BAYESIAN PRIORS 60

E.5 COMPUTATIONAL METHODS 61 E.5.1 Population Monte Carlo 61 E.5.2 Hybrid MCMC methods 61 E.5.3 Computational of mixture distribution 61 E.5.4 Importance sampling 61

E.6 BAYESIAN ESTIMATION Of QUANTILE DISTRIBUTIONS 62

E.7 DNA SEGMENTATION ANALYSIS 63

E.8 ANALYSING AND REPORTING CLINICAL INDICATORS 65 USING BAYESIAN HIERARCHICAL MODELS

E.9 BAYESIAN HIDDEN MARKOV MODEL fOR DNA SEQUENCES 65 SEGMENTATION MODELLING

E.10 SMOOTH TEST Of GOODNESS Of fIT 66

E.11 HIERARCHICAL MODELLING 66

F. DISTRIBUTED SENSING AND CONTROL 67

f.1 NETWORKED CONTROL 67 f.1.1 Analysis and design of networked control systems 67

f.2 POWER TRANSfORMER MODELLING 68 f.2.1 Distributed power transformer model for partial discharge location 68 f.2.2 The influence of inductive disparity on the observed frequency 69 response of a distribution transformer f.2.3 The influence of complex permeability on the broadband 70 frequency response of a power transformer

f.3 ROBOTICS RESEARCH 71 f.3.1 Automated colour calibration system using multivariate gaussian 71 mixtures to segment colour space f.3.2 Sound-scapes for robot localisation through dimensionality reduction 71 f.3.3 A low power walk for the NAO robot 72

f.4 SYSTEM IDENTIfICATION TOOLS fOR NEURON RESPONSE EVENTS 72

G. MATHEMATICAL SYSTEMS THEORY 73

G.1 BEYOND THE SPECTRUM 73

G.2 NONLINEAR ANALYSIS, OPTIMISATION AND fIXED-POINT THEORY 74 G.2.1 Alternating-projection and reflection algorithms 75 G.2.2 Analysis in the absence of linearity 75 G.2.3 fixed points in the absence of weak compactness 75 G.2.4 Semigroups of mappings 75 G.2.5 Ultraproduct methods 75

PUBLICATIONS 76

BOOKS 76

BOOKS IN PREPARATION/TO APPEAR 76

CHAPTERS IN BOOKS 76

PATENTS 76

PLENARY AND KEYNOTE ADDRESSES 76

INVITED PRESENTATIONS 77

JOURNAL PAPERS 77

JOURNAL PAPERS ACCEPTED fOR PUBLICATION 80

CONfERENCE PAPERS 81

TECHNICAL REPORTS TO INDUSTRY 84

PERFORMANCE INDICATORS REPORT 86

INCOME AND EXPENDITURE STATEMENT 92

Page 5: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control04

OUR VISION

To be a world leader in analysis, design and optimisation of complex dynamic systems; pursuing outstanding fundamental and applied research.

Page 6: 2008 - University of Newcastle

2008 ANNUAL REPORT 05

DIRECTOR’S REPORT

Welcome to the 2008 Annual Report for the ARC Centre of Excellence for Complex Dynamic Systems and Control. The report outlines our achievements in 2008 including both theoretical results and practical results of relevance to industry.

As in the 2007 report we will include all publications written by members of the Centre. This is intended to give a better overall view of the depth of research work within the Centre.

Also, as in the 2007 report, we will include an outcomes section. This was suggested by the ARC as a mechanism for giving greater prominence to the part of our work which is linked to industrial research and development.

On the topic of outcomes, in 2008 the ARC held an Outcomes forum to highlight practical outcomes of research funded by the Australian Government. An associated book describes 31 meritorious projects. Members of CDSC were honoured to have three contributions appear in the book, namely;

Tristan Perez, “Optimising the performance of marine vessels”

Greg Adams, “Making advances in mineral exploration”

Graham Goodwin, “Advanced control: surmounting complexity”

These contributions highlight work done with our industrial partners and affiliates. This work is a particular focus of our research activities. Indeed, we currently have over 20 separate industrial research and development projects with 11 national and international companies.

On a personal note, I was honoured to receive the IfAC Quazza Medal. This medal is one of the highest awards from IfAC (The International federation of Automatic Control). I was delighted to receive the award not only as a personal achievement but also because the award indirectly recognizes the wonderful team within which I have the privilege to work.

We believe that CDSC provides an exciting and vibrant research environment spanning basic research through to relevant industrial applications. The work would not be possible without the support of many organisations. In particular, we wish to gratefully acknowledge financial support from the Australian Research Council, NSW Department of State and Regional Development, The University of Newcastle, Queensland University of Technology, our industrial partners, and our industrial affiliates.

finally, I would like to take this opportunity to express my thanks to the full CDSC team for their hard work and support during the past year.

Graham C. GoodwinDirector and Laureate Professor

Page 7: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control06

STAff STAff MOVEMENTS

Director

Laureate Professor Graham C. Goodwin

Associate Directors

Professor Minyue fu

Professor S.O. Reza Moheimani

Chief Operating Officer

Dr. Greg Adams

Program Leaders

Dr. Julio Braslavsky Industrial Control and Optimisation

Professor Minyue fu Signal Processing

Professor Kerrie Mengersen Bayesian Learning (Statistics, QUT Node)

Professor Reza Moheimani Mechatronics

Professor John Rayner Statistical Inference and Modelling (Statistics, UoN Node)

Dr. Maria Seron Control System Design

Associate Professor Brailey Sims Mathematical Systems Theory

Dr. James Welsh Distributed Sensing and Control

Other Investigators

Dr. Jose De Doná

Professor Anthony Pettitt (QUT)

Dr. Robert Reeves (QUT)

Dr. Ian Turner (QUT)

Dr. George Willis

Industrial Liaison Officer

Dr. Tristan Perez

Industry Partner Investigators

Dr. Salvatore (Sam) Crisafulli (Matrikon)

Dr. Merab Menabde (BHP-Billiton Innovation)

Dr. James B. Lee (BHP-Billiton Innovation)

Dr. R. Arthur M. Maddever (BHP-Billiton Innovation)

Mr. Peter Stone (BHP-Billiton Innovation)

Mr. Richard Thomas (Matrikon)

Industrial Affiliates

Boeing Research and Technology, Australia

CfW Hamilton Jet

Connell Wagner

CSR Sugar

Halcyon International

Hatch IAS

CDSC Funded Researchers

Dr. David Allingham

Dr. Sumeet Aphale

Dr. Bharath Bhikkaji

Dr. Andrew fleming

Dr. Mark Griffin (QUT)

Dr. Peter Howley

Dr. Katrina Lau

Dr. Paula Lennon (QUT)

Dr. Ross McVinish (QUT)

Dr. Kaushik Mahata

Dr. Damian Marelli

Dr. Adrian Medioli

Mr. Sean Moyniham (QUT)

Dr. Claus Müller

Dr. Jun Ning

Dr. Darfiana Nur

Dr. Jaime Peters (QUT)

Dr. Alejandro Rojas

Dr. Elizabeth Stojanovski

Dr. frank Tuyl

Dr. Ian Wood (QUT)

Dr. Trent Yeend

Dr. Yuenkuan Yong

Dr. Mei Mei Zhang

Dr. Jinchuan Zheng

Associated Researcher

Professor Rick Middleton

Engineering Staff

Mr. frank Sobora

Support Staff

Mrs. Dianne Piefke

Mrs. Jayne Disney

n Jun Ning commenced a one year contract in June 2008, working in the Signal Processing programme.

n Bharath Bikkaji, Sumeet Aphale, Carlos Ocampo-Martinez, Wojciech Szymanski and Ross McVinish all moved to other institutions during the year.

Page 8: 2008 - University of Newcastle

2008 ANNUAL REPORT 07

POSTGRADUATE RESEARCH STUDENTS (CONTINUING)

Brendan Burke (Commenced in 2008)Thesis Title: “Constrained multi variable control of an integrated sugar mill system for economic enhancement”Supervisor: G.J. AdamsCo-Supervisor: G.C. GoodwinDegree: ME

Francisco Eduardo Castillo SantosThesis Title: “Aspect of metric fixed point theory”Supervisor: B. SimsCo-Supervisor: G. WillisDegree: PhD

Ben DeanThesis title: “Modelling with the generalised lambda distribution”Supervisor: R. KingCo-Supervisor: P. HowleyDegree: MPhil

Kingsley EzehThesis title: “Statistical techniques for improving health care”Supervisor: P. HowleyCo-Supervisor: R. KingDegree: MSc

Paul FaheyThesis title: “Analysis of performance data for universities and hospitals”Supervisor: P. HowleyCo-Supervisor: J. RaynerDegree: PhD

Naomi HendersonThesis title: “Improving Robot Vision using Spatial and Temporal Correlations”Supervisor: R. KingCo-Supervisor: S. Chalup/R.H. MiddletonDegree: PhD

Kenny HongThesis title: “face perception and face expression dynamics”Supervisor: S. ChalupCo-Supervisor: R. King/R.H. MiddletonDegree: PhD

Wenbiao Hu (QUT)Thesis Title: “Bayesian spatio-temporal CART”Supervisor: K. MengersenDegree: ME

Sandra Johnson (QUT)Thesis Title: “Bayesian networks”Supervisor: K. MengersenDegree: PhD

Jason Kulk Thesis Title: “Anthropomorphic stance and walk for consumer robots”Supervisor: J.S. WelshCo-Supervisor: Z. ChenDegree: PhD

John Philip LaneThesis title: “Delivery of social services in an restructing region”Supervisor: R. KingCo-Supervisor: E. StojanovskiDegree: PhD

Kate Lee (QUT)Thesis Title: “MCMC algorithms”Supervisor: K. MengersenCo-Supervisor: R. McVinishDegree: PhD

Iskandar MahmoudThesis Title: “High-speed scanning probe microscopy”Supervisor: S.O.R. MoheimaniCo-Supervisor: B.M. NinnessDegree: PhD

Steven MitchellThesis Title: “Physical interpretation of the broadband frequency response of power transformers”Supervisor: J.S. WelshCo-Supervisor: R.E. BetzDegree: PhD

Steven NicklinThesis Title: “Using model predictive control in bipedal locomotion”Supervisor: J.S. Welsh/R.H. MiddletonCo-Supervisor: K. MahataDegree: PhD

Chris Oldmeadow (QUT)Thesis Title: “Bayesian latent variable models”Supervisor: K. MengersenDegree: PhD

Paul Rippon Thesis title: “Application of smooth tests of goodness of fit to generalised linear models”Supervisor: J. Rayner Co-Supervisor: K. Mengersen Degree: PhD

Ian SearstonThesis Title: “Analysis in geodesic metric spaces”Supervisor: B. SimsCo-Supervisor: G. WillisDegree: PhD

Aurelio Salton (Commenced in 2008)Thesis Title: “Dual-stage control”Supervisor: M. fuCo-Supervisor: Zhiyong Chen

Degree: PhD

Matthew SkerrittThesis Title: “Recent developments in fixed point theory”Supervisor: B. SimsCo-Supervisor: J. BorweinDegree: MPhil

Fatimah Almah SaaidThesis title: “Demand forecast with time series data mining approach”Supervisor: D. NurCo-Supervisor: R. KingDegree: PhD

Mark SmithThesis Title: “Ultramethods in metric fixed point theory”Supervisor: B. SimsCo-Supervisor: G. WillisDegree: PhD

Fajar SuryawanThesis Title: “Nonlinear model predictive control”Supervisor: J.A. De DonaCo-Supervisor M.M. SeronDegree: PhD

Lukasz WiklendtThesis Title: “Learning and control in robotics”Supervisor: S. ChalupCo-Supervisor: R.H. MiddletonDegree: PhD

Page 9: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control08

Aaron WongThesis Title: “Sound-scape visualisation through dimensionality reductionSupervisor: S.ChalCo-Supervisor: K.MahataDegree: PhD

Alain Yetendje LemegniThesis Title: “fault tolerant multisensory systems”Supervisor: M.M. SeronCo-Supervisor: J.A. De Doná Degree: ME

THESES SUBMITTED IN 2008

Milan DerpichThesis Title: “Sampling and quantisation in audio compression”Supervisor: G.C. Goodwin Co-Supervisor: D. QuevedoDegree: PhD

Wang MengThesis Title: “Parsimonious information structures in real time signal processing”Supervisor: G.C. Goodwin Co-Supervisor: D. QuevedoDegree: PhD

Eduardo SilvaThesis Title: “Performance limitations in networked control”Supervisor: G.C. Goodwin Co-Supervisor: D. Quevedo

Degree: PhD

Darren Wraith (QUT)Thesis Title: “Bayesian mixture models for environmental health”Supervisor: K. MengersenDegree: PhD

Zhuo, Xiang WeiThesis Title: “Estimation and control: Multisensors, explicit solutions and duality”Supervisor: J.A. De DonáDegree: PhD

GRADUATED 2008

Boris GodoyThesis Title: “Modelling and control of copper leaching processes”Supervisor: J.H. BraslavskyCo-Supervisor: R.H. MiddletonDegree: PhD

Christian LøvaasThesis Title: “Robust MPC”Supervisor: G.C. GoodwinCo-Supervisor: M.M. SeronDegree: PhD

Jose MareThesis Title: “Constrained tracking estimation: Analytical solutions, symmetry and nonlinear insights”Supervisor: J.A. De DonáCo-Supervisor: G.C. Goodwin Degree: PhD

Adrian MedioliThesis Title: “Constraints, stability and feasibility issues in model predictive control”Supervisor: M.M. Seron Co-Supervisor: R.H. MiddletonDegree: PhD

Trevor MoffietThesis Title: “Statistical methods for software and decision report for remote sensing analysis”Supervisor: R.King/K. MengersenDegree: PhD

Cristian RojasThesis Title: “Robust experiment design” Supervisor: J.S. Welsh Co-Supervisor: G.C. GoodwinDegree: PhD

L to R. Boris Godoy, Graham Goodwin and Adrian Medioli on Graduation Day, 10 October 2008.

Page 10: 2008 - University of Newcastle

2008 ANNUAL REPORT 09

ADVISORY BOARD

Professor R. JarvisDepartment of Electrical and Computer Systems Engineering, Monash University, Melbourne, Victoria

Professor J. CarterPro. Vice-Chancellor, Faculty of Engineering and Built Environment, The University of Newcastle, Callaghan, NSW

Professor I.M.Y. MareelsMelbourne University, Melbourne, Victoria

Mr. R. PeirceTechnical Systems, CSR Victoria Mill, Ingham, Queensland

Professor I.R. PetersenAustralian Defence Force Academy, UNSW, Canberra, ACT

Mr. Colin ThewNSW Department of State and Regional Development, Sydney, NSW

Professor A. SharmaDeputy Vice-Chancellor, Queensland University of Technology,

Dr. Brent JenkinsChief Executive Officer, TUNRA Limited, Callaghan, NSW

CHAIRMAN

Professor Barney Glover, Deputy Vice-Chancellor, Research, The University of Newcastle

CURRENT MEMBERS

Distinguished Professor Brian D.O. AndersonResearch School of Information Sciences and Engineering, The Australian National University, Canberra, ACT

Professor A. CareyMathematical Sciences, Australian National University, Canberra, ACT

Dr. S. CrisafulliMatrikon, Warrabrook, NSW

Dr. W.J. EdwardsHatch IAS, Teralba, NSW

Dr. S. GaleaCSIRO, DSTO, Melbourne, Victoria

Mr. R. HayesShell Refining (Australia) Pty. Ltd., Clyde Refinery, Rosehill, NSW

Professor W. HogarthPro. Vice-Chancellor, Faculty of Science and Information Technology, The University of Newcastle, Callaghan, NSW

The CDSC Advisory Board met in Newcastle on Friday 25 July

2008 to review progress, consider management issues and offer

advice on strategic directions for the Centre. Professor Pat Michie,

Pro Vice-Chancellor, Research chaired the meeting.

Page 11: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control10

Dr Damien FrancoisUniversité Catholique de Louvain, Louvaine-la-Neuve, Belgium October

Professor Peter GawthropCentre for Systems and Control, Glasgow University, Scotland June

Professor Jan Tommy GravdahlDepartment of Engineering Cybernetics, Norwegian University of Science and Technology, Trondheim, Norway July 2007 – July 2008

Dr. Kazimarez Goebel Institute of Mathematics, Marie-Curie Sklodowska University, Lublin, Poland february – March

Dr Hernan HaimovichDepartmento de Electronica, Universidad Nacional de Rosatio, Argentina. December 2008 – March 2009

Associate Professor Renquan LuInstitute of Information and Control, Hangzhou Dianzi University, Hangzhou, P.R. China August – November

Professor Christian RobertDepartment of Statistics, Université Paris Dauphine, france August

ACADEMIC VISITORS

Professor Mazen AlamirCNRS, GIPSA-lab, Control Systems Department, St Martin d’Hères, france January – September

Dr Karina BarbosaNational Laboratory for Scientific Computing, Rio de Janeiro, Brazil August 2007 – May 2008

Professor Daniel CoutinhoPontifıcia Universidade Catolica do Rio Grande do Sul, Brazil february

Dr Stephen DuncanDepartment of Engineering Science, University of Oxford, United Kingdom September

Dr Alicia Esparza PeidroDepartment Ingenieria de Sistemas Y Automatica, Universidad Politecnica de Valencia, Spain June – July

Professor Antoine Ferreira ENSI, Bourges, france february

Professor Arie FeuerDepartment of Electrical Engineering, Technion – Israel Institute of Technology, Haifa, Israel. July – October

VISITORS

Page 12: 2008 - University of Newcastle

2008 ANNUAL REPORT 11

Dr Monica E RomeroDepartamento de Electronica, Universidad Nacional de Rosario, Argentina february – April

Dr Oliver ThasDepartment of Applied Mathematics, Biometrics and Process Control, Ghent University, Belgium february

Professor Vincent WertzCESAME, Université Catholique de Louvain, Louvaine-la-Neuve, Belgium August 2007 – January 2008

Professor Robert WolpertDepartment of Statistical Science, Duke University, North Carolina, USA July

Professor Lihua XieDivision of Automation and Control, Nanyang Technological University, Singapore. August 2007 – January 2008

Dr Juan I. Yuz E Departamento de Electrónica, Universidad Técnica federico Santa Maria, Valparaiso, Chile July – August

STUDENT VISITORS

Ms Diana UgryumovaUniversity of Twente, The Netherlands March – August

Mr. Emiliano Pereira GonzalezUniversity of Castilla, La Mancha, Spain January – June

Mr. Elias HerreroUniversity of Cantabria, Spain March – October

Mr. Tahar SalmaPRISME Institute, ENSI Bourges, france January – March

Mr Tai, XinNational Laboratory of Industrial control Technology, Zhijiang University, Hangzhou, Peoples Republic of China. November 2007 – April 2009

Mr Alain Yetendje LemegniEcole Polytechnique, University of Marseille, france. September 2007 – January 2008

Dr Maria Seron (CDSC) and Dr Monica Romero (Argentina) collaborated on research related to fault detection in electrical machines.

Page 13: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control12

INDUSTRY INTERACTION AND SELECTED OUTCOMES 2008

One of the prime goals of CDSC is to contribute to the development of industrial competitiveness and capabilities of our industry partners and affiliates through the application of innovative research, and the delivery of specialised industrial courses. CDSC provides a flexible scheme for industry interaction with three modes: consultancy, an affiliate’s program, and a partner program. Each of these modes targets different needs of support, contribution and interaction with industry.

In 2008, CfW Hamilton Jet & Co. (New Zealand) and Boeing Research & Technology Australia became industrial affiliates of the centre. CDCS also delivered a one-week course on modelling, identification and control, and conducted 20 research projects in collaboration with 11 industrial partners from Australia, New Zealand, Norway, and Spain – some of the project descriptions can be found in the “Industrial Control and Optimisation” research programme.

for further details on how to establish industrial collaborations with CDSC, contact the centre Industry Liaison Officer, Dr. Tristan Perez, [email protected]

Support from the NSW Department of State and Regional Development. We greatly appreciate and acknowledge the support of the NSW Department of State and Regional Development (DSRD), who contribute significant funds to CDSC each year for the purposes of developing NSW industries and research capabilities.

In 2008, our DSRD funds were used in the following ways:

n Substantial support for the Sferics project (BHP-Billiton Innovation, Newcastle)

n Development of the Virtual Laboratories, including the funding of a stand at the 17th IfAC World Congress (Seoul, South Korea, July 2008)

n Work on the WestVAMP cogeneration project (Illawarra Coal, Appin, NSW);

n funding of the CDSC Industrial Development Scholarship to the amount of $6,500 per year for 3 years. This scholarship was awarded to Mr. Hal Cooper, who is developing virtual laboratories for high school education.

Selected OutcomesThe interactions with industrial partners aims to produce significant outcomes for each industry. In 2008, selected industrial partners have benefitted through the following projects.

BHP-Billiton: Sferics ProjectThe partnership with CDSC has continued to benefit BHP-Billiton in 2008 in the development of leading edge mineral exploration technology. One of the principal technologies for the discovery of many types of mineral deposits, in particular those associated with conductive mineralisation such as nickel and silver/lead/zinc, is time-domain electromagnetic (TEM) surveying. A key factor in the improvement of TEM surveying technology is the reduction of sferics, i.e. environmental noise originating in lightning discharges around the globe.

Since 2007, the Sferics project has developed techniques that assist in the removal of sferic noise. In 2008 these techniques have been extended to the

removal of broadband sferics noise between 4 Hz and 1 kHz. A collection of software routines that implement the sferic removal algorithms were written for BHP-Billiton. Independently, CDSC has also contributed to the improvement of other aspects of TEM surveying technology by studying new sensors models, which can be used to compensate the sensor transient response. These advancements offer the potential to increase the depth of TEM surveying, thereby opening up new ground to exploration by this method.

Dr. R. Arthur M. MaddeverPrincipal Scientist Resource R&D Resource Business Optimisation BHP-Billiton

BHP-Billiton: Integrated Mine PlanningIn 2008, BHP Billiton’s collaboration as an Industrial Partner in the CDSC has led to several outcomes of substantial benefit in improving our understanding of operational and market risk. Additionally, we have closely collaborated with CDSC to develop extensions to planning tools and algorithms which have improved our ability to optimise both strategic mine plans and corresponding infrastructure investment decisions.

The major focus of our research collaboration has been in the area of so-called 2-factor stochastic price modelling of mineral commodities. In this research area, we have achieved several milestones, which together form the foundation of a world-class in-house capability in the identification, validation and deployment of advanced forward commodity price models – a capability that allows BHP Billiton a real strategic advantage in optimising the strategic management of a large project portfolio. These milestones include:

n Explanation and validation of incomplete published two-factor stochastic price model identification algorithms.

n Development of faster and more robust identification algorithms, which are verifiably superior to any in the published literature.

Page 14: 2008 - University of Newcastle

2008 ANNUAL REPORT 13

n Development of methods to validate the stationarity of the identified model parameters and to nominate which parameters are material and significant.

n Application of the identification algorithms to real commodity data streams (both spot and forward historical price data) for aluminium, copper and oil.

The resulting stochastic price estimation models are a critical input into BHP Billiton’s models for understanding the full value of our resource assets, specifically in recognizing the value of future embedded management options which are a direct outcome of the level of equilibrium price uncertainty. In ongoing research, we hope to develop robust methods for extending the parameter identification to a multi-commodity case.

In the area of strategic mine plan optimisation technology development, CDSC has developed a world-first mining phase optimisation heuristic which has been demonstrated to improve net present value of existing phase designs by the order of 5%, whilst enforcing difficult practical mining access constraints. We currently have plans in train to embed this technology within our proprietary mine planning software tools.

finally, we have been for a short while collaborating with CDSC in an investigation into the reliability and accuracy of Discrete Event Simulation (DES) models in informing important business decisions, particularly around multi-billion dollar mine capacity expansions or greenfield developments. The focus of this research is to understand how different DES modelling strategies can lead to different business decisions, and furthermore to identify more robust model parameterizations – particularly of system disturbances – which will reduce the chance of DES model pathologies affecting the business decisions.

It follows from the above, the outcomes of CDCS in 2008 in relation to BHP Billiton projects have been most successful, and we look forward to continuing our collaboration with CDSC.

Peter M. StoneSenior Principal ScientistBusiness Optimisation TeamBHP Billiton

Halcyon International Pty. Ltd.: Improved Ride Control of Marine VesselsThe initial collaboration between Halcyon and CDSC, which initiated in 2007, lead to an enhancement of Halcyon’s commercial ride control system. This system is currently in use on the patrol boats operated by the Australian Customs Service, and a large fast ferry operating across the straits of Gibraltar. With this work, Halcyon also won the 2007 Western Australia Award of Excellence in Engineering for Small Business given by The Institute of Engineers Australia.

In 2008, we continued the work on ride control, and CDSC assisted Halcyon with the development of numerical tools for design and performance prediction of ship roll gyrostabilisers (devices that use gyroscopic forces of spinning masses to attenuate the ship roll motion induced by the waves). In particular, two computer codes were developed: GYRODESIGN and GYROSIM. GYRODESIGN is a frequency-domain analysis code that uses a model of the gyros, vessel hydrodynamic data, and vessel mass distribution to obtain the response operators of roll motion for the vessel with and without the gyro-stabiliser. This tool allows the designer to size the gyros for a particular vessel by providing initial performance assessment. The code also provides information necessary for mechanical and hydraulic designs. GYROSIM is a tool that performs batch-case simulations in time domain for specific irregular sea states and sailing conditions. Once the initial sizing of the gyros is done with GYRODESIGN, GYROSIM provides a refinement of the predicted performance. GYROSIM uses comprehensive mathematical models and also control algorithms that aim at maximising the performance of the gyro in changing sea states via a combination of parameter estimation and control adaptation.

The software developed in collaboration with CDSC has a significant value for Halcyon since it aids the design process and reduces design time, and it also contributes to improve the performance prediction of Halcyon gyrostabilisers. Several aspects of the project have been identified for further development in 2009.

Paul SteinmannManaging DirectorHalcyon International Pty LtdLevel 1, 6 – 8 Mews Roadfremantle, Western Australia

CSR Sugar: Economic Enhancement of an Integrated Sugar MillThe co-generation plant at CSR Sugar’s Pioneer Mill (Ayr, NQ) has brought significant benefits to CSR’s business, notably in efficiency improvements, carbon abatement and new income streams. The main factors in running the co-generation plant involve the more efficient operation of process units, and the optimal sharing of mill resources between units, in particular steam.

Multi-effect evaporators are key process units in a sugar mill, as it is there where the most efficient evaporation of water from sugar syrup occurs. Pioneer Mill currently suffers from unsatisfactory control of sugar content (brix) in syrup from the evaporator sets, resulting in less efficient use of steam, and hence loss of export power generation capability.

In 2008, CSR Sugar’s partnership with CDSC has resulted in notable improvements to brix control, and the identification of other possible improvements to be tested in the 2009 crushing season. Looking further ahead, it is expected that our work together can optimise operations at Pioneer Mill significantly through better steam usage.

Rob PeirceManager, Technical SystemsCSR SugarIngham, QLD

Page 15: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control14

A number of courses and workshops were organised during the year. Details are:

n CDSC Short Courses on Modern Industrial Control (2 days – attended by 10); System Modelling and Identification (1.5 days – attended by 12) and Kalman filtering and Applications (1.5 days – attended by 7) ran from Monday 11 february until friday 15 february. Presenters of this short course were: Graham Goodwin, Minyue fu, Julio Braslavsky, Tristan Perez, James Welsh, Greg Adams, and Daniel Quevedo, with Greg Adams and Alejandro Rojas running the Laboratories.

n Kerrie Mengersen arranged the International Bayesian and Statistics Conference held on Hamilton Island: A Workshop on Complex Computer Models held in Sydney; a 5 day short course on Bayesian Analysis which was held in Brisbane; and “Bayesian Core”, a 5 day short course held in Brisbane and presented by Prof Christian Robert from Université Paris, Dauphine, france.

n Kerrie Mengersen presented a 4 day workshop in Brisbane entitled “Bayesian Statistics for Beginners”.

n The 2008 Summer Systems and Control Workshop, jointly organised by Reza Moheimani, School of Electrical Engineering and Computer Science, University of Newcastle, and The School of Electrical and Telecommunications Engineering, University of New South Wales, was held at The University of Newcastle on friday 15 february 2008. This workshop was attended by approximately 30 students. Presenters were: Professor Antoine ferreira, ENSI Bourges, france; Professor Ian Petersen, UNSW@ADfA; Dr. Elanor Huntington, UNSW@ADfA; Dr. Andrew fleming, The University of Newcastle.

n James Welsh arranged a Postgraduate Student Signals and Systems Workshop which took place at the University of Newcastle on 17 and 18 April 2008. The Workshop was attended by students and staff from the Australian

Defence force Academy, Australian National University, University of Melbourne, University of New South Wales and The University of Newcastle. A total of 23 presentations were given by postgraduate students in areas covering control applications, nonlinear control, networked control, quantisation, applications in optimisation, system identification and estimation.

n Tristan Perez organised an internal Workshop on “Large Scale Problems within Integrated Mine Planning”. Presenters were Rick Middleton, MeiMei Zhang (BHP-Billiton Melbourne) and Claus Mueller.

Participants of the Postgraduate Student Signals and Systems Workshop held in Newcastle

CONfERENCES, COURSES AND WORKSHOPS 2008

Page 16: 2008 - University of Newcastle

2008 ANNUAL REPORT 15

SEMINARS 2008

16 January 2008Author: Professor Rick Middleton The Hamilton Institute, NUI, Maynooth, Co Kildare, Ireland.Title: Performance limitations in control of distributed systems with limited communications

16 January 2008Author: Steven Nicklin School of Electrical Engineering and Computer Science, the University of Newcastle.Title: Biped locomotion using MPC

30 January 2008Author: Dr John K. Ward, CSIRO Energy Technology Australia.Title: Outside the box – HVAC management techniques that consider the world around them

4 February 2008Author: Iskandar Mahmood School of Electrical Engineering and Computer Science, the University of Newcastle.Title: High-speed scanning probe microscopy

14 February 2008Author: Dr Eric Beh School of Computing and Mathematics, The University of Western Sydney.Title: Exploring ways of measuring association in contingency tables evaluation

27 February 2008 Author: Naomi Henderson School of Electrical Engineering and Computer Science, the University of Newcastle.Title: Image classification using multivariate Gaussian mixtures: Applications in robotics and medical imaging

12 March 2008 Author: Dr Michael Lundh ABB, Sweden.Title: Machine direction control

28 March 2008Author: Dr David Allingham CDSC, The University of Newcastle.Title: The Mechanics of ABC, or, Say, Isn’t That...?

23 April 2008Author: Dr Robert King School of Mathematical and Physical Sciences, The University of NewcastleTitle: Partial classification of non-unique objects with rpart for robot soccer

23 June 2008Author: Professor Peter Ramadge, Department of Electrical Engineering, Princeton University, USA.Title: Advanced signal processing problems in functional MRI analysis

27 June 2008Author Dr John Best School of Mathematical and Physical Sciences, the University of NewcastleTitle: Tests of fit for the Beta-Binomial Distribution

27 June 2008Author: Dr Darfiana Nur School of Mathematical and Physical Sciences, the University of NewcastleTitle: Hidden Markov Models for DNA sequence segmentation: Simulation and evaluation

11 August 2008Author: fajar Suryawan, School of Electrical Engineering and Computer Science, The University of Newcastle.Title: Study of the use of flatness based control, splines parameterization, and MPC for reference trajectory tracking in nonlinear systems

Research students and staff from the CDSC, as well as Australian and international visitors participate in the Centre’s seminar series. Seminars presented in 2008 are listed as follows:

Page 17: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control16

26 August 2008Author: Jason Kulk, School of Electrical Engineering and Computer Science, The University of Newcastle.Title: Anthropomorphic stance and walk for consumer robots

3 September 2008Author: Professor Nick Longford Universitat Pompeu fabra, Spain Title: Small-area estimation with spatial similarity

17 September 2008Author: Associate Professor Yan-Xia Lin University of Wollongong

Title: A graphical method for identifying potential domains for change points in generalised bernoulli processes

26 September 2008Author: Prof John Rayner School of Mathematical and Physical Sciences, The University of NewcastleTitle: Testing equality of variances for multiple univariate populations

25 September 2008 Author: Professor Arie feuer, faculty of Electrical Engineering, Technion, Israel.Title: Blind multi-band signal reconstruction

1 October 2008 Author: Professor Reza Moheimani School of Electrical Engineering and Computer Science, The University of Newcastle.Title: Control issues in nanopositioning

3 October 2008Authors: Dr Robert King, Dr Liz Stojanovski, Ms Maureen Townley-Jones School of Mathematical and Physical Sciences, The University of NewcastleTitle: Attracting postgrads, engaging undergrads, and encouraging reflection: Reflections on Ozcots 08

15 October 2008Author: Dr George Sofronov School of Mathematics and Applied Statistics, University of Wollongong Title: Statistical model-based optimisation

23 October 2008Authors: Megan ford, Steven Kennedy, Paul Rippon, Almah Saaid, Klairung Pearl Samart, Ian Robinson, Kenny Hong, Philip Lane and Naomi Henderson The University of NewcastleTitle: Annual Postgraduate Presentation Day

24 October 2008Author: Dr frank Tuyl Hunter New England Population Health. Title: The Rule of Three, its variants and extensions

17 November 2008 Author: Associate Professor Steve Keen, School of Economics and finance, University of Western Sydney.Title: Dynamic modeling of economic systems and financial instability

Page 18: 2008 - University of Newcastle

2008 ANNUAL REPORT 17

SELECTED HIGHLIGHTS 2008

n Graham Goodwin was presented with the Quazza Medal at the 17th IfAC World Congress in Seoul, South Korea.

n CDSC held its annual retreat in October. The day was attended by all Centre academic staff, general staff, postgraduate students, visitors, as well as representatives from industry.

n Minyue fu presented a Plenary Address at the Deans Conference on Education of Control and Automation, China, in April. The title of his address was “Education of Control and Automation: Challenges and future”.

n Numanoids (NUBOTS) won the 2008 Standard Platform Soccer World Cup in China.

n The paper, “Design, Analysis and Control of a fast Nanopositioning Stage” co-authored by Yuen Yong, Sumeet Aphale and Reza Moheimani was selected as a finalist paper for the 2008 IEEE/ASME Advanced Intelligent Mechatronics Conference Best Paper Award.

n Dr. Jose Mare (PhD student supervised by Jose De Doná) received, in 2008, the annual faculty Award for Research Higher Degree Excellence based on the nomination and comments made by the panel of international examiners on the quality of his PhD thesis.

n Jose De Doná received the 2008 Pro Vice-Chancellors Award for Research Excellence. These awards aim to encourage and recognise the achievements of early and mid-career researchers. Research excellence is judged relative to career opportunity and candidates are judged by several criteria including: originality of the research; quality of the research output; the impact of research; success in attracting research funds; success in research higher degree supervision; peer recognition; and track record relative to opportunity.

n Tristan Perez was appointed “Adjunct Associate Professor of Ship Dynamics” at the Norwegian University of Science and Technology (NTNU), Trondheim, Norway.

Presentation ceremony where Graham Goodwin received the Quazza Medal

n Kerrie Mengersen was Program Chair for the 2008 International Society of Bayesian Analysis Conference, which was held for the first time in the Australasian region, on Hamilton Island, Queensland. Dr Clair Alston was Chair of the Organising Committee for this meeting.

n The ISI Essential Science Indicator has listed five papers from Centre researchers in the top 1% of Highly Cited papers for the last 10 years – Moheimani (with fleming, Aphale and Bhikkaji – 3), Borwein (1) and Szymanski (1).

Page 19: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control18

RESEARCH PROGRAMS

08

Page 20: 2008 - University of Newcastle

2008 ANNUAL REPORT 19

Program Goals:

The partnerships between researchers and industry enable reciprocal transfer of knowledge and new ideas of great potential impact on the community and economy. This program encompasses several research projects motivated by and in collaboration with industrial partners. The main underlying theme of these projects is the application of advanced modelling, control and optimisation techniques to maximise asset utilisation and improve performance. The complexity of the dynamics of such processes arise from factors including model errors, unknown disturbances, nonlinearities, distributed parameter systems, elements of Human-Machine Interaction and hybrid (discrete and continuous state) components. Expected outcomes of the program include high quality research solutions and human resources tailored to the needs of Australian industry.

In 2008 we welcomed CfW Hamilton Jet & Co (New Zealand) and Boeing Research & Technology, Australia as Industrial Affiliates, joining Halcyon International, CSR, Connell Wagner and Industrial Automation Services (Hatch IAS).

A.1 PERfORMANCE OPTIMISATION Of MARINE SYSTEMS

Project Leader: T. Perez

Researchers: C. Løvaas, G.C. Goodwin and J.C. Agüero

External Academic Collaborators: T.I. fossen (Norwegian University of Science and Technology, Norway) C. Holden (Norwegian University of Science and Technology, Norway) E. Herrero, (Student, University of Cantabria, Spain)

External Industrial Collaborators:P. Steinmann (Halcyon International, Australia)T. Armstrong (Austal Ships, Australia)J. Borret (Hamilton Jet, New Zealand)M. Santos-Mujica (Robotiker-Tecnalia, Spain)S. Peder-Berge (Offshore Simulator Centre, Norway)

Marine systems are designed to perform complex operations that require appropriate reliability and economy. These requirements demand an interdisciplinary approach to address the tight integration of design aspects related to hydrodynamics, structures, and motion control.

This project is dedicated to the design of tools for guidance and motion control with the aim of optimising the performance of marine vehicles in different operations. The project targets vessels and operations within offshore, maritime transport, underwater exploration, unmanned vehicles and wave energy conversion. Some of the current research is being conducted together with international academic and industry collaborators.

A. INDUSTRIAL CONTROL AND OPTIMISATION

Julio BraslavskyProgram Leader

Tristan PerezDeputy Program Leader

A

figure 1: Marine systems are designed to perform complex operations that require appropriate reliability and economy. These requirements

demand an interdisciplinary approach to address the tight integration of design aspects related to

hydrodynamics, structures, and motion control. (Picture courtesy of Offshore Simulator Centre AS, Norway).

Page 21: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control20

−200

2040

60−40

−200

2040

−30

−20

−10

0

10

Y−axis (m)

3D Visualisation of the Wamit file: semisubow.gdf

X−axis (m)

Z−ax

is (m

)A.1.1 Marine system simulation tools

Researchers: T. Perez and T. fossen (Norway)

The Marine Systems Simulator (MSS) is an environment developed to provide the necessary resources for rapid implementation of mathematical models of marine systems with focus on control system design. The platform adopted for the development of the MSS is Matlab/ Simulink. This allows a modular simulator structure, and the possibility of distributed development. Openness and modularity of software components have been prioritised in the design. This enables a systematic reuse of knowledge and results in efficient tools for research and education.

In 2008, we added functions for frequency-domain identification, which are an aid to the construction of time-domain models from frequency-domain data computed from hydrodynamic codes. The latest version of the software and future updates can be freely downloaded from www.marinecontrol.org

A.1.2 Identification of radiation force parametric models of marine structures from 2D frequency domain data

Researchers: T. Perez and T. fossen (Norway)

The ability to predict ship responses and loads in waves is an important tool in the design of marine structures and motion control systems. One method for constructing time-domain models consists of using the data generated by the hydrodynamic codes to compute the different elements of the so called Cummins’ equation of ship motion. In this project, we have been studying the application of different identification methods in both time and frequency domain to make best use of the available hydrodynamic data and constraints based on prior knowledge derived from hydrodynamic theory.

In 2008, we addressed the problem of joint identification of infinite-frequency added mass and fluid-memory models of marine structures from finite frequency data. This problem is relevant for cases where the code used to compute the hydrodynamic coefficients of the marine structure does not give the infinite frequency added mass. This case is typical of codes based on 2D-potential theory codes. The method proposed presents a simpler alternative approach to other methods previously presented in the literature, and the same identification procedure can be used to identify the fluid-memory models with or without having access to the infinite-frequency added mass coefficient. Therefore, it provides an extension that puts the two identification problems into the same framework.

10−2 10−1 100 10 190

100

110

120

130

140

150

Freq. [rad/s]

|K(jw

)|

Convolution Model DoF 33

10−2 10−1 100 10 1−100

−50

0

50

100

Phas

e K(

jw) [

deg]

Freq. [rad/s]

10−2 10−1 100 1010

0.5

1

1.5

2

2.5x 107

Freq. [rad/s]

B [K

g/s]

Potential Damping DoF 33

10−2 10−1 100 1014.5

5

5.5

6

6.5

7

7.5x 107

Freq. [rad/s]

A [K

g]

Added Mass DoF 33

K(jw)Khat(jw) order 8

BBest FD ident, order 8

AAest FD indet, order 8Ainf

A

a) b)

figure 2: frequency domain identification of marine structure dynamics base on 2D hydrodynamic data. (a) shows the geometry of an offshore rig used as an application example (data from www.marinecontrol.org). (b) shows the identification results for the frequency response of the fluid memory function in heave based on a parametric model of order 8 without using infinite frequency added mass as part of the available data. This figure also shows the reconstruction of added mass and damping from the identified parametric model.

Page 22: 2008 - University of Newcastle

2008 ANNUAL REPORT 21

AA.1.3 Modelling and control of parametric resonance in marine vessels

Researchers: T. Perez, T. fossen (Norway) and C. Holden (Norway)

Parametric resonance is a phenomenon where changes in the model parameters can be used to describe rapid build-up of oscillations. This phenomenon has been observed in the rolling motion of contemporary ship designs with significant bow flare and raised sterns. The resonance can be developed when sailing in a head and stern seas with a wave length similar to the length of the ship. In these conditions, the hydrodynamic forces that restore the up-right equilibrium of the vessel present a time-varying characteristic, which depends on the location of the wave crest along the length of the vessel. This is equivalent to a mass-spring-damper system where the stiffness of the spring varies with time. This effect results in roll parametric resonance due to changes in the restoring forces, which creates a rapid development of roll motion that can reach up to 40deg in just a few roll cycles. This phenomenon has been responsible for containers being washed overboard in container ships and capsizing fishing vessels.

This project examines the modelling of this phenomenon and the use of controlled roll stabilisation devices to reduce the effect on the vessel motion. In 2008, a parametric model previously proposed in the literature has been fitted to experimental data of a scale model of a 300m containership, and a model of the coupled ship and an actively controlled u-tank stabiliser has been developed.

A.1.4 Experiment design and identification of marine vessels dynamics

Researchers: T. Perez, G.C. Goodwin, J.C. Agüero, E. Herrero (Spain) and T. Armstrong (Austal Ships)

The mathematical models of marine vessels can be obtained from first principle or analytical modelling. The parameters of the model, however, often need to be estimated from full scale trials. The ability of an estimation method to produce good estimates of the parameters depends on how much information about the dynamics of the system is contained in the data, which in turn depends on the experiment performed. Hence, the design of experiments is of paramount importance to obtain accurate model parameters and to reduce the time to perform experiments. Some of the hydrodynamic phenomena involved in ship dynamic response are too complex for analytical modelling. Therefore, parts of the model structure may also be determined based on analysis of experimental data.

This project examines the design of optimal experiments and the application of system identification or experimental modelling for vessels performing different operations: dynamic positioning, manoeuvring at low speed and manoeuvring at high speed.

In 2008, we focused on a two-stage approach for experiment design, in which we first collect data from step responses, and then based on the information gathered, appropriate signals for parametric model identification are designed. This method has been applied to the identification of vessels for positioning and slow speed manoeuvring using data of full scale trials of a small fishing vessel. In addition, we have proposed the application of statistical methods to select hydrodynamic damping model structures that provides best use of the information available. These methods have been applied to simplify the manoeuvring model of a 130m a novel fast ferry trimaran designed and built by Austal Ships, Australia.

0 50 100 150 200 250 300 350

u

TrialModel

0 50 100 150 200 250 300 350Time [s]

v

EstimModel

0 50 100 150 200 250 300 350

p

TrialModel

0 50 100 150 200 250 300 350Time [s]

r

TrialModel

a)

b)

figure 3: Time-domain identification of a coupled 4DOf manoeuvring model of a highspeed vessel. (a) shows Austal’s High-Speed Trimaran Design (Picture courtesy of Austal Ships, Australia). (b) shows the identification results of a manoeuvring parametric model with structure selected via stepwise regression in the time domain. The figure shows the model and fullscaletrial responses in velocity: surge (u), sway (v), roll (p), and yaw (r).

Page 23: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control22

AA.1.5 System identification for rapid model prototyping of ship training simulators (Offshore Simulator Centre AS, Norway.)

Researchers: T. Perez and S. Peder-Berge (Norway)

Ship training simulators are used to improve crew efficiency and thus safety of marine operations. At the core of any virtual-reality simulator lays a mathematical model that describes the ship dynamic response to control and environmental forces. When a new vessel is to be incorporated into a simulator, different type of data for the vessel may be available to be used for system identification to extract a mathematical model for dynamic response.

In 2008, CDSC evaluated various identification methods for obtaining parametric models of vessel response based on frequency-domain data computed by hydrodynamic codes, and made recommendations to the Offshore Simulator Centre AS as to which method to use for rapid identification of vessel dynamic response.

figure 4: Ship training simulators are used to improve crew efficiency and thus safety of marine operations. At the core of any virtual-reality simulator lays a mathematical model that describes the ship dynamic response to control and environmental forces. When a new vessel is to be incorporated into a simulator, different type of data of the vessel may be available to be used for system identification to extract a mathematical model for dynamic response.

Page 24: 2008 - University of Newcastle

2008 ANNUAL REPORT 23

AA.1.6 Adaptive control of gyroscopes for roll stabilisation of marine vessels (Halcyon International, Australia)

Researchers: T. Perez and P. Steinmann (Halcyon)

The use of gyroscopic effects of high speed spinning masses for the roll stabilisation of marine structures was proposed over 100 years ago. This approach was very effective, but limited control and large sizes hindered further developments. In recent years, there has been significant interest in revitalising gyrostabilisers due to improvements in materials, bearings, and lubricants, which have contributed to fast spinning devices and size reduction. In addition, fast and

reliable computer control systems ensure adequate operation over an envelope of sailing conditions.

This project has examined methods for determining the size of the gyros to achieve a desired level of roll reduction and also for control design to ensure performance in a range of sailing conditions. A time-domain simulation tool GYROSIM has been developed, which allows a rapid evaluation of the expected performance of gyrostabilisers.

In 2008, a frequency-domain simulation tool GYRODESIGN has been developed for a rapid gyro sizing and to provide design information related to mechanical and hydraulic component specification. With this

tool, a designer is able to have a fast initial assessment of the design before refining the performance predictions using the time-domain simulation code GYROSIM.

Also in 2008, an adaptive control strategy was designed to maximize performance of the gyrostabiliser in changing environmental conditions. This strategy aims at dealing with a fundamental trade-off in gyrostabiliser control design in which high performance in roll reduction must be balanced with potential constraint violations when a large group of waves arrive at the vessel. The results have been tested in simulation and an experimental prototype is being built for further testing in real sea conditions.

figure 5: Halcyon’s twin gyrostabiliser. A gyrostabiliser consists of one or more spinning wheels whose gyroscopic-induced forces are used to counteract the forces induced by the waves on a ship, and thus reduce motion. In a roll gyrostabiliser, the spinning wheel is positioned such that the gyroscopic effect reduced the motion in roll. The use of twin wheels rotating in opposite directions eliminates the gyroscopic effects in other degrees of freedom than the one intended to be controlled.

figure 6: Numerical study of roll gyrostabilisation of a navy patrol boat using Haylcyon’s gyro stabilisers. The top plot shows the roll angle at zero forward speed in a 3m sea state as a function of the wave period. The bottom plot shows the expected percentage of roll angle reduction (RMS).

Page 25: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control24

AA.1.7 Control design for optimum power extraction in wave energy converters (Robotiker-Tecnalia, Spain)

Researchers: T. Perez and M. Santos-Mujica (Spain)

The search for renewable energy resources has revitalised the interest in devices for wave energy conversion. Wave energy converters (WEC) extract energy from the motion induced by the waves on particular hull designs. In order to maximize the extracted energy, the design of a control below to regulate the loads of the power take off element is of great importance.

In this project, we look at the control aspects of WEC. In particular, we are studying fundamental performance limitations that can affect potential control system designs for maximum energy extraction. The preliminary results are exciting and we thus plan to submit an ARC Discovery Project dedicated to this topic in 2009.

In 2008, we have analysed a particular WEC device, which uses a large gyroscope mounted on a floating platform. This device extracts energy from precession motion induced on the gyroscope as a result of the pitching motion of the structure. Based on a model obtained from a combination of hydrodynamic computations and experimental tests, we have computed an upper bound on the expected power to be extracted in a various sea states. We also have guided ROBOTIKER in the development of a time-domain simulation package, and designed an adaptive precession torque control that aims at optimising the extracted power.

0

50

100

150

200

250

300

350Power Extraction−Head Seas

W

Pa/ζ

2 [kW

/m2 ]

Ideal PTOGyro Bg = 15, 35, 65 x sqrt(4 Ig Cg)

a) b)

A.1.8 Constrained predictive control of ship fin stabilizers to prevent dynamic stall

Researchers: T. Perez and G.C. Goodwin

In moderate to high sea states, the effectiveness of ship fin stabilisers can severely deteriorate due to nonlinear effects arising from unsteady hydrodynamic characteristics of the fins: dynamic stall. These nonlinear effects take the form of a hysteresis, and they become very significant when the effective angle of attack of the fins exceeds a certain threshold angle. Dynamic stall can result in a complete loss of control action depending on how much the fins exceed the threshold angle. When this is detected, it is common to reduce the gain of the controller that commands the fins. This approach is cautious and tends to reduce performance when the conditions leading to dynamic stall disappear. An alternative approach for preventing the effects while keeping high performance consists of estimating the effective angle of attack and set a conservative constraint on it as part of the control objectives. In our work, we have investigated the latter approach, and here proposed the use of a model predictive control (MPC) to prevent the development of these nonlinear effects by considering constraints on both the mechanical angle of the fins and the effective angle of attack.

figure 7: Performance analysis of a wave energy conversion device with different control designs. figure (a) shows an schematic of a particular wave energy converter that uses a power take off element (PTO) to extract energy from the wave-induced pitch motion. figure (b) shows the estimated power extraction per unit of wave amplitude as a function of the wave frequency. The ideal PTO is an upper bound on performance estimated using only hydrodynamic characteristics. The use of different control designs and a particular PTO was used to evaluate the potential gains of control adaptation to changes in the dominant wave frequency.

Page 26: 2008 - University of Newcastle

2008 ANNUAL REPORT 25

AA.2 OPTIMISATION BASED OPERATOR GUIDANCE SCHEMES

Project Leader: J.H. Braslavsky

Researchers: G.J. Adams, J.-C. Agüero G.C. Goodwin, B. Godoy (Student), and A.J. Rojas

External Academic Collaborators:J.T. Gravdahl (Norwegian University of Science and Technology, Norway), D. Ugryumova (Student, University of Twente, The Netherlands)

External Industrial Collaborators:D. Boggs (BHP-Billiton, Perth, Australia)M. Downey (BHP-Billiton, Newcastle, Australia)J. Lee (BHP-Billiton, Newcastle, Australia) A. Maddever (BHP-Billiton, Perth, Australia)R. Turner (BHP-Billiton, Newcastle, Australia)

This project is funded by a partnership of the Centre with BHP Billiton, and deals with the development of new technologies using mathematical modelling, and state-of-the-art model-based control and estimation tools. The project currently encompasses three sub-projects:

n Modelling and control of copper heap bioleaching processes,

n Sferics reduction in electromagnetic mineral exploration,

n Cogeneration at WestVAMP

A.2.1 Modelling and control of copper heap bioleaching processes

Researchers: J.H. Braslavsky and B. Godoy

This sub-project focuses on the development of mathematical models and control strategies for heap bioleaching processes for the extraction of copper from sulphide minerals. Heap bioleaching is of increasing interest in the mining industry to recover metals from secondary ores. See CDSC Annual Reports 2003-2007 for more background information.

In previous work, we have suggested the use of feedback control to improve the rate of mineral extraction based on linearized models around nominal trajectories of the output of interest. In 2008, this work has been refined in a comparative study between two feedback approaches: Model Predictive Control (MPC) and Extremum Seeking Control (ESC). Previously obtained linearized models were used to design an MPC strategy incorporating input constraints. ESC was tuned, without requiring a process model, to maximise copper extraction rate using aeration rate.

Simulation results show that similar copper extraction rates can be obtained using either strategy. The extraction rates improvements with respect to an optimised fixed set-point strategy are between 4-5%, which correspond to 840-1,050 extra tonnes a year for a small bioleaching facility. Both feedback strategies improve robustness with respect to model uncertainties. A paper communicating these results has been presented in the 17th IfAC World Congress in Seoul, South Korea.

This project has been the topic of research of Boris Godoy for his PhD thesis, which was awarded in October 2008.

Page 27: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control26

+ −

Manipulatedsetpoints

Process

++ u y

∆ y∆u

y0outputs

Nominalsetpointsu0

Estimated nominal

MPCLinear

ModelBHPB

Manipulatedsetpoints

Sustainedexcitation

Process

uk yk

×LPF+

HPF

− γz − 1

A sin(ω k )

A

figure 10: Comparison of extraction rates improvements for MPC and ESC with respect to an optimised fixed setpoint strategy. MPC requires less control efforts, while ESC presents slightly faster recovery.

figure 8: MPC application to heap bioleaching. A high complexity model is used to estimate linearised models around nominal trajectories. MPC is tuned to track increments in heap average temperature by manipulating aeration rate and raffinate influx. In a real implementation the model could be retuned adaptively.

figure 9: ESC feedback strategy schematics. Adaptation is applied on the aeration rate setpoint based on measurements of copper concentration is leached solution.

Page 28: 2008 - University of Newcastle

2008 ANNUAL REPORT 27

AA.2.2 Sferics reduction in electromagnetic mineral exploration

Researchers: J.-C. Agüero, J.H. Braslavsky, M. Downey, G.C. Goodwin, K. Lau, J.B. Lee, A. Maddever, P. Turner and D. Ugryumova

This is a joint project with BHP-Billiton Exploration and Mining in Perth (previously in Newcastle).

The aim of this industry project is the reduction of sferics noise in mineral exploration using Geoferret, an Australian designed and developed electromagnetic exploration system. The exploration technique employed by Geoferret relies on the induction of currents in the earth followed by the measurement of the magnetic field generated by the induced currents. The reduction of sferics noise, electromagnetic noise originating from lightning storms, is central to the improvement of signal to noise ratio for the detection of deeper ore bodies. See CDSC Annual Report 2006 for more background information.

In 2007, broadband noise reduction was achieved using separate models for the low and high frequency ranges. Progress in 2008 has included the development of a single, unbiased model for broadband (4 Hz-1 kHz) multinode noise cancellation. A Matlab implementation of the model estimation and noise cancellation algorithms has also been written for BHP-Billiton. Other work includes the estimation of a model for new sensors. This model can be used for compensation of the sensor response.

In March, Diana Ugryumova, a Masters Student from the University of Twente in The Netherlands, visited for six months to complete the practical component of her course. During her visit she worked on the application of our modelling techniques to impedance estimation in another electromagnetic exploration method: magnetotelluric sounding.

In July, a paper on the application of errors-in-variables techniques to sferics attenuation was presented at the IfAC World Congress 2008 in Seoul, Korea. The technique is used to estimate a model which is deployed for noise cancellation. The estimated model agrees well with

those obtained using alternative methods, with similar performance on experimental data. The results indicate that the technique can be successfully applied in this case and shows potential for other applications. See Lau, Braslavsky, Aguero and Goodwin (2008) in Conference Papers.

future work is planned for the extension of our noise cancellation techniques to other types of magnetic field sensors currently being tested by BHP-Billiton.

A.2.3 Co-generation at WestVAMP

Researchers: G.J. Adams, G.C. Goodwin, J.T. Gravdahl (Norway) and A.J. Rojas

This project is aimed at improvements to the control for the Westcliff Vent Air Methane Project facility near Appin, NSW. The details are subject to a confidentiality agreement, but overall the facility aims to generate power from the methane that is present in air vented from underground coal mines. Work in 2008 focussed on deriving a model of the process, and control improvements are being investigated using this model.

100 101 102 103 10410−14

10−13

10−12

10−11

10−10

10−9

10−8

10−7

f (Hz)

PSD

(V2 /H

z)

Noise cancellation results − PSD

Zresidual

100 101 102 103 1040

0.1

0.2

0.3

0.4Noise cancellation results − Coh ZX

100 101 102 103 1040

0.2

0.4

0.6

0.8Coh ZY

f (Hz)

Zresidual

figure 12: Noise cancellation results. Coherence between the Z and X components and the Z and Y components of the sferics noise before and after performing noise cancellation. The coherence has been reduced to a negligible level at most frequencies between 4 Hz and 1 kHz. This indicates that almost all of the correlated noise has been eliminated.

figure 11: Noise cancellation results. Power spectral density of the measured sferics noise Z and the residual after performing noise cancellation. The 50 Hz powerline harmonics have not been removed.

Page 29: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control28

AA.3 INTEGRATED MINE PLANNING (BHP BILLITON)

Project Leader: T. Perez

Researchers: G.C. Goodwin, M. fu, M.M. Zhang, B. Godoy (Student), X. Tai (Student)

External Academic Collaborators: K. Barbosa (National Laboratory for Scientific Computing, Rio de Janeiro, Brazil)

External Industrial Collaborators:P.M. Stone (BHP-Billiton) M. Menabde (BHP-Billiton)

The integrated mine planning project aims at developing tools for optimisation of planning and operation of mines. In 2008, CDSC worked on three particular aspects of the problem. The first aspect is parameter estimation of models for commodity prices. The second aspect concerns the development of planning tools and algorithms to optimise both strategic mine plans and infrastructure investment decisions. The third aspect is the optimal dispatching scheduling of truck/trains within a mine.

The major focus of commodity price model estimation was to develop algorithms for parameter estimation. Commodity price modelling is normally approached in terms of structural time-series models, in which the different components (states) have a clear financial interpretation. In particular, we looked at the parameter and state estimation of two-factor models. These models use a trend towards an equilibrium price (long term component) and a reversion to such a trend (short term component), which represents the difference between the current commodity price and the equilibrium price.

We have developed two different algorithms based on the Maximum-likelihood approach. The first algorithm uses a full parameterisation, and thus results in a non-linear parameter optimisation problem. This algorithm provides high quality estimates provided that a good initial guess of the value of the parameters is available. The second algorithm solves a sequence of estimation problems that increase in complexity and

uses the estimates of one stage to initialise the subsequent stage. This algorithm is, thus, self-initialised, and the resulting estimates also give the option of initialising the first proposed algorithm for a further refinement of the parameter values. figure 13 shows a particular realisation of a simulated metal price and future contracts. figure 14 shows the estimates of the main parameters of the model corresponding to 100 different realisations of simulated data using the self-initialising algorithm.

The resulting price models are a critical input into BHP Billiton’s models for understanding the full value of resource assets and for recognizing the value of future embedded management options. The latter are a direct outcome of the level of equilibrium price uncertainty. In an ongoing research robust methods for extending the parameter identification to a multi-commodity case are being researched.

The second aspect of mine planning looked into in 2008 was mining phase design, which is a critical step in long-term mine planning process and heavily affects the net present value of a life-of-mine plan. In BHP Billiton’s proprietary mine planning tool, the mine planning process are divided into a few sub processes which are optimised separately. The mining phases are developed from an “optimal” block extraction sequence from a prior “block aggregates” scheduling using a fuzzy clustering algorithm. The mining phases developed in this way sometimes are not practical, and manual post-processing is required to ensure practical for mining operation. furthermore, the existing mining phase design method does not optimise the net present value.

To address these issues, we have developed a heuristic automatic method to enforce the practical mining access constraints. We have embedded the method into a framework of meta-heuristic optimisation. Testing has demonstrated that the new approach can improve the net present value of existing phase designs by the order of 5%.

Also some preliminary investigation has been carried out on the problem of optimal train/truck dispatch using approximate dynamic programming and discrete event dynamic simulation.

Page 30: 2008 - University of Newcastle

2008 ANNUAL REPORT 29

0 50 100 150 200 250

2

2.5

3Simulated Data

Time [Weeks]

Spot

Pric

e

0 50 100 150 200 250

0.7

0.8

0.9

1

1.1

Time [Weeks]

Log

Spot

Pric

e

Spot1 month5 months9 months13 months17 months24 months

0 10 20 30 40 50 60 70 80 90 10012

14

16

18

20Intitial Estimates Usign Spot & Future Prices

Realization

Kapp

a

0 10 20 30 40 50 60 70 80 90 100−0.1

0

0.1

0.2

Realization

ξ

0 10 20 30 40 50 60 70 80 90 1000.160.18

0.20.220.24

Realization

χ 0

0 10 20 30 40 50 60 70 80 90 1000.68

0.7

0.72

Realization

ξ 0

figure 13: Sample realisation of a commodity spot and future contract prices based on simulated

data. The data is generated using a model with parameters estimated from market data of the

London Metal Exchange.

figure 14: Prediction error parameter estimation results over 100 simulated realizations of commodity

prices and future contracts. This figure shows the estimates and true values of the main parameters of

a 2-factor commodity price model. The parameters are the strength of mean reversion (Kappa), trend (mu), and the short and long term initial conditions of the states of the model. The analysis based on

simulated data is used to evaluate the properties of the estimation algorithm; that is, it is used to gain

confidence in the results with real data.

A

Page 31: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control30

0 50 100 150 200 250 300 350 4000

0.5

1

1.5

Out

put a

mpl

itude

r1y1 (original)y1 (decoupled)

0 50 100 150 200 250 300 350 400−0.5

0

0.5

1

1.5

Out

put a

mpl

itude

time (sec)

r2y2 (original)y2 (decoupled)

0 50 100 150 200 250 300 350 400−0.2

0

0.2

0.4

Inpu

t am

plitu

de

0 50 100 150 200 250 300 350 400−0.2

−0.1

0

0.1

0.2

Inpu

t am

plitu

de

time (sec)

u1 (original)u1 (decoupled)

u2 (original)u2 (decoupled)

figure 16: The associated MV moves in the presence of MV constraints.

figure 15: Coupled (green) and decoupled (blue) setpoint change responses in the presence of MV constraints.

A.4 NEXT GENERATION MODEL BASED CONTROL TOOLS (MATRIKON)

Project Leader: G.J. Adams

Researchers: N. Germyn (Student), G.C. Goodwin, A.M. Medioli, R.H. Middleton, M.M. Seron and J.S. Welsh

External Academic Collaborators:D. francois (Université Catholique de Louvain, Belgium)

External Industrial Collaborators:P. farragher (Matrikon) R. Thomas (Matrikon)

A.4.1 Next generation model based control tools for CPO

Researchers: G.J. Adams, N. Germyn, G.C. Goodwin, A.M. Medioli, R.H. Middleton, M.M. Seron, R. Thomas and J.S. Welsh

The aim of this project is to deliver to Matrikon process control tools that allow

n appropriate handling of complex, nonlinear and heterogeneous processes;

n robust and easy-to-use system identification; and

n economic optimisation of process variables.

Progress in 2008 occurred in the following areas:

n Tools for economic optimisation, including data reconciliation strategies and interfaces for steady-state optimisation of linear and non-linear objectives, were developed and integrated into CPO (this was completed as an honours student project by Nic Germyn).

n Code changes in the CPOmpc tool have resulted in major execution speed improvements being attained, and setup modifications allow on-line model changes to occur smoothly.

n Handling of non-linear processes via the appropriate scheduling of multiple linear controllers was made possible by the development of an MPC Scheduler.

n Matrikon are developing a CPOmpc control solution platform with an overseas company, which will involve the control of non-linear systems via multiple linear regions.

n Decoupling strategies that are intrinsic to the internal QP objective function of the CPOmpc tool were integrated and tested. Some results are shown in figure 15 and figure 16.

future work on next-generation model-based control tools for CPO will investigate:

n Derivation of MPC tuning parameters from parameters of existing (non-MPC) control.

n Robust MPC, based on the algorithm of Løvaas, Seron and Goodwin (see Journal Publication)

n Testing of the CPOmpc scheduling tool applied to non-linear processes.

n Extending and integrating the economic optimisation features as part of a “dynamic process optimisation” loop; some extensions include strategies for extremal seeking, and checks for infeasible setpoints.

n A major case study is planned for 2009, which will bring together the features of non-linear multivariable control, system identification and economic optimisation into the control and optimisation of a simulated nutating grinding mill.

A

Page 32: 2008 - University of Newcastle

2008 ANNUAL REPORT 31

1 2 3 4 5 6 70

10

20

30

40

50

60

70

80

90

100

Number of Predictions

Cor

rect

Cla

ssifi

catio

n Pe

rcen

tage

Batch CentroidAdaptive CentroidStatic Centroid

A.4.2 Next generation model based control tools for ProcessMORE

Researchers: P. farragher, D. francois (Belgium) and A.M. Medioli

This project involves the development of algorithms for automatically generating downtime causes from alarm sets. CDSC work in 2008 involved:

n A thorough investigation of alarm data in relation to physical plant layout and components.

n A study of possible or expected classification performance, investigating issues such as new alarms and downtimes being added over time, irrelevant alarms etc.

n A visit by Damien francois (Belgium), who developed a simple classifier in 2007.

n Modification and enhancement of Damien’s classifier, and its application to new data.

n Moving towards a system that may be used on-line, to adapt to changing alarm sets and plant configuration.

The performance of the original “batch” and static classifier and the latest classifier is summarised in figure 17, where the “Adaptive Centroid” method would be most useful in a real (on-line) system; results show that classification performance for the Adaptive Centroid method, when the system suggests up to seven possible causes, has over 50% accuracy.

Items to work on in 2009 include:

n Strategies for culling irrelevant data.

n Applying the system to other alarm data.

n Different handling of data pre and post-downtime.

n Better algorithms to try to improve performance.

figure 17: Comparison of classifier performance for different centroid generation methods.

A

Page 33: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control32

A.5 CSR SUGAR (INDUSTRIAL AffILIATE)

Project Leader: G.J. Adams

Researchers: B.J. Burke (Student, CSR Sugar), G.C. Goodwin, A. Rayner (Student), A.J. Rojas and B. Sims

External Academic Collaborator:J.T. Gravdahl (Norwegian University of Science and Technology, Norway)

External Industrial Collaborator: R.D. Peirce (CSR Sugar)

A.5.1 Constrained, multi-variable control of an integrated sugar mill system for economic enhancement

Researchers: G. Adams, B. Burke, G. Goodwin, J. Gravdahl (Norway), R. Peirce, A. Rojas

In today’s carbon – and energy-conscious economy, manufacturers are searching for ways to maximise returns and minimise wastage. This is certainly true in CSR Sugar, where at Pioneer Mill (near Ayr, North Queensland) a co-generation plant has been installed. This plant uses waste cane fibre (bagasse) from a number of mills to create steam for both sugar milling and for electricity generation. If steam is used for sugar milling efficiently, CSR can export 50MW of power to the local grid, gaining an income stream from a waste product.

This project aims to study energy and steam use in sugar processing at Pioneer. The multi-effect evaporators are core units in the process, where the most efficient evaporation of water from syrup occurs. Proper control of the sugar content (brix) coming out of the evaporators, and coordination of steam/energy use with other sections of the mill, are essential to minimise energy losses.

The current control of the brix suffers from periodic disturbances (90 minute periods), as well as quicker oscillatory disturbances (10 minute period) caused by addition of water to the syrup upstream during mill stoppages. CSR and CDSC have studied evaporator operation, and have determined the cause of the quicker oscillation to be a type of non-linear flow reduction through the outlet valves. With strategies in place to reduce this non-linear flow effect in evaporator “5A”, oscillations are reduced from those seen in figure 18 to those in figure 19. Evaporator “5B” still suffers from these oscillations in figure 19.

Investigation continues into the cause of the slower (90 minute) oscillations. These oscillations are most likely to be due to the steam demand from the batch crystallization pans. Signal analysis is being done on selected process variables to uncover significant correlations.

A

Page 34: 2008 - University of Newcastle

2008 ANNUAL REPORT 33

A.5.2 CSR brake van control

Researchers: A. Rayner and B. Sims

Sugar cane trains use a wagon at the end of the train, called a brake van, to control train braking. The main aim of the brake van is to keep the couplings between each of the cane bins in tension. Once the couplings go into compression, derailments can occur, especially when the bins are empty. The operation of the brake van is via radio link from the locomotive, and consists of a numbered dial with increasing amounts of brake pressure. A park brake can also be independently applied.

CSR is looking to improve the control of the brake van with an automated process (i.e. the brake van automatically selecting the appropriate level of braking for the conditions), as well as implementing a new braking unit. In making these improvements, CSR is looking to increase the efficiency of the whole process of transporting cane. By automating the braking system they will effectively save money on driver training, reduce fuel needs (since the brakes are being used more effectively) and cut down on the number of replacements to the brake pads. Replacing the current brake type with an electrical braking system will save further, since electrical systems brake much faster than the current system.

Work performed by CDSC on this project so far involves initial modelling of the forces involved in the carriages/couplings. further work will concentrate on this modelling aspect, and it is envisaged that systems which incorporate GPS information as well may be useful.

figure 19: Improved brix control in 5A (red) compared with 5B (blue).figure 18: Examples of poor brix control behaviour in evaporators 5A (red) and 5B (blue).

A

Page 35: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control34

A.6 CONNELL WAGNER (INDUSTRIAL AffILIATE)

Project Leader: J.S. Welsh

Researchers: D. Allingham and J.S. Welsh

External Industrial Collaborators:J. Tusek (Connell Wagner)

This project, being undertaken in conjunction with Connell Wagner, is investigating methods of parameter estimation for synchronous machines.

Its aim is to provide validation for methods already in use, to explore alternative approaches and to investigate ways of providing error estimates for parameter values.

Synchronous machines are the primary generators of electricity for the power production industry. Estimation of machine parameters is a vital field of study, with literature dating back to the 1920s. Many approaches are available, using different measurement and model regimes. Most current practices are based upon recent IEEE standards (for example, IEEE Std 115-1995 and IEEE Std 1110-2002) which describe in detail both standstill and on-line tests for synchronous machines.

Our work to date has focussed on standstill frequency response (SSfR) modelling. Here, sinusoidal inputs over a range of frequencies, approximately from 0.5 mHz to 1 kHz, are applied to the machine in a variety of configurations, and the machine response is measured. Using these responses, transfer functions are calculated and the parameters of an equivalent circuit model of the machine, shown in the accompanying figure, are then estimated. The measured responses along with fits obtained from two estimation methods are also shown.

The parameters from the SSfR tests will, in turn, be used to estimate the machine response for on-line step tests, which involve small changes to the amplitude of the machine’s power when it is under load (for example, when it is connected to the power grid). from these tests, time constants are estimated which describe the machine’s response to events such as faults, and how fast it can recover from such events.

A.7 HATCH (INDUSTRIAL AUTOMATION CONTROL) (INDUSTRIAL AffILIATE)

Project Leader: G.C. Goodwin

Researchers: A. Rojas, C. Renton (Student), G.C. Goodwin

External Industrial Collaborators:T. Domanti (Hatch IAS) G. Wallace (Hatch IAS)

In 2008, the work focussed on cross directional control issues in galvanizing lines. The project had two streams:

(i) Rapid estimation of coating thickness under non-stationary conditions. Here it was found that it was desirable to use an estimator having variable memory. (A conference paper has been written based on this work).

(ii) Combining strip location with double sided measurements to give improved thickness estimation.

The latter was principally the work of C. Renton who did this as part of his final year honours project.

A.8 BOEING RESEARCH AND TECHNOLOGY, AUSTRALIA (INDUSTRIAL AffILIATE)

Project Leader: J.S. Welsh

Researchers: T. Perez and J.S. Welsh

External Industrial Collaborators: B.P. Williams (Boeing Research and Technology, Australia) V. Wheway (Boeing Research and Technology, Australia)

This project began in December 2008. The details of the project are confidential, but involve the development of flight systems for autonomous aircraft.

1

adL

kdr

kdl

kdr

2

21

fkd1l fkd2lar ll fdl fdr

kdl

dV fdV

fdIdI

10−3

10−2

10−1Zdd

Gai

n

10−4 10−2 100 1020

30

60

90

Phas

e−sh

ift

10−4

10−3

10−2

10−1

100Zfd

10−4 10−2 100 1020

30

60

9010−3

10−2

10−1Zdd0

10−4 10−2 100 1020

30

60

9010−3

10−2

10−1pGfd

10−4 10−2 100 102−90

−60

−30

0

30

60

90

10−2

10−1

100

101Zff

Gai

n

10−4 10−2 100 1020

30

60

90

Phas

e−sh

ift

10−4

10−3

10−2

10−1Zdf

10−4 10−2 100 1020

30

60

9010−2

10−1

100

101Zff0

10−4 10−2 100 1020

30

60

9010−1

100

101pHdf

10−4 10−2 100 102−90

−60

−30

0

30

60

90

10−3

10−2

10−1Zdd

Gai

n

10−4 10−2 100 1020

30

60

90

Phas

e−sh

ift

10−4

10−3

10−2

10−1

100Zfd

10−4 10−2 100 1020

30

60

9010−3

10−2

10−1Zdd0

10−4 10−2 100 1020

30

60

9010−3

10−2

10−1pGfd

10−4 10−2 100 102−90

−60

−30

0

30

60

90

10−2

10−1

100

101Zff

Gai

n

10−4 10−2 100 1020

30

60

90

Phas

e−sh

ift

10−4

10−3

10−2

10−1Zdf

10−4 10−2 100 1020

30

60

9010−2

10−1

100

101Zff0

10−4 10−2 100 1020

30

60

9010−1

100

101pHdf

10−4 10−2 100 102−90

−60

−30

0

30

60

90

figure 21: Plots of transfer function estimates vs. real data for a large synchronous machine.

figure 20

A

Page 36: 2008 - University of Newcastle

2008 ANNUAL REPORT 35

Program Goals:

This program is aimed at developing advanced control design methods to tackle challenging problems associated with modern and emerging mechatronic systems. Over the past few years, particular attention has been paid to micro – and nanoscale mechatronic systems. More specifically, to develop methodologies, technologies and the necessary mechatronic instrumentation for fast and accurate interrogation and manipulation of matter at the nanoscale. This research is performed by our multidisciplinary research team and at the Laboratory for Dynamics and Control of Nanosystems (LDCN), a multi-million dollar state-of-the-art research facility dedicated to the advancement of nanotechnology through innovations in systems theory and control engineering.

B.1 CHARGE DRIVES fOR SCANNING PROBE MICROSCOPE POSITIONING STAGES

Project Leader: A.J. fleming

External Academic Collaborator: K.K. Leang (Virginia Commonwealth University, USA)

A key component of scanning probe microscopes is the piezoelectric nanopositioning system required to manoeuvre the probe or sample. Due to hysteresis exhibited by piezoelectric actuators, positioning stages in scanning probe microscopes require sensor-based closed-loop control. Although closed-loop control is effective at eliminating nonlinearity at low scan speeds, the bandwidth compared to open loop is severely reduced. In addition, sensor noise significantly degrades achievable resolution.

In this project, charge drives were developed as a simple alternative to closed-loop control when feedback cannot be applied or provides inadequate performance. Such situations arise in high-speed imaging, where position sensor noise can be large or where no feedback sensors are present. Charge drives were shown to reduce the error caused by hysteresis to less than 1% of the scan range. These improvements were demonstrated in the first known experimental images using charge drive, shown in figure 22.

B. MECHATRONICS

S.O. Reza Moheimani Program Leader

A.J. fleming Deputy Program Leader

figure 22: A comparison of 50x50um images recorded using voltage (left) and charge actuation (right). The sample is a periodic calibration grating with 20nm feature height.

B

Page 37: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control36

BB.2 SIMULATION Of DYNAMICS COUPLING IN PIEZOELECTRIC TUBE SCANNERS BY REDUCED ORDER fINITE ELEMENT MODELS

Project Leader: A.J. fleming

External Academic Collaborators: J. Maess (University of Stuttgart, Germany) f. Allgöwer (University of Stuttgart, Germany)

Piezoelectric tube scanners are thin cylinders of piezoelectric material, widely used in scanning probe microscopes to position the sample or probe. fast and accurate scanning requires the suppression of dominant low-frequency resonances as well as the compensation of dynamics-coupling effects. This project aims to procure detailed finite element models of the strain deformation in piezoelectric tube scanners. A goal was to explain the cross-coupling between axes and the existence of unexpected high-frequency modes. Such dynamics are not explained by present models but are evident in experimental results.

The finite element package ANSYS was used with customized elements developed to represent piezoelectric behaviour. The resulting simulations compared well with an experimental modal analysis performed using a PSV300 Laser Scanning Vibrometer. Previously unexplained resonance modes were attributed to circumferencial bending modes and a low-frequency torsional mode. The simulated and experimentally measured deformation of the first lateral bending mode is illustrated in figure 23.

figure 23: The simulated (left) and experimentally measured (right) deformation of the first lateral bending mode.

Page 38: 2008 - University of Newcastle

2008 ANNUAL REPORT 37

B.3 MINIMIZING SCANNING ERRORS IN PIEZOELECTRIC STACK-ACTUATED NANOPOSITIONING PLATfORMS

Project Leader: S.O.R. Moheimani

Researchers: S. Aphale, B. Bhikkaji and S.O.R. Moheimani

Many types of nanopositioning devices are constructed using piezoelectric stack actuators and compliant flexures. The piezoelectric actuators provide atomic resolution, and the flexures eliminate out-of-plane or rotational travel. Unfortunately, due to the low stiffness of compliant flexures, the mechanical resonance frequency of a typical stack actuated nanopositioner is only a few hundred Hertz. This results in a low-frequency lightly damped resonance that severely limits closed-loop performance.

This project proposes and compares three types of damping controller to attenuate the first mechanical resonance and allow increased closed-loop bandwidth. A polynomial-based pole placement controller is shown to provide the best performance while being simple to design. Experimental results demonstrate a closed-loop resolution of approximately 8 nm, over a range of 0.1mm

B.4 HIGH-BANDWIDTH CONTROL Of A PIEZOELECTRIC NANOPOSITIONING STAGE IN THE PRESENCE Of PLANT UNCERTAINTIES

Project Leader: S.O.R. Moheimani

Researchers: S. Aphale and S.O.R. Moheimani

External Academic Collaborators: S. Devasia (University of Washington, USA)

Inversion-based feedforward techniques can deliver accurate tracking performance in the absence of plant parameter uncertainties. Piezoelectric stack actuated nanopositioning platforms are prone to variations in their system parameters, especially the resonance frequencies which are dependent on temperature and the load mass.

In this project, a new technique was proposed and tested for high-performance tracking control of a nanopositiong system that is insensitive to changes in resonance frequency. The proposed technique combines a feedback controller that reduces the impact of parameter uncertainty, with an inversion-based feedforward technique. The combination is shown experimentally to increase the tracking bandwidth of a nanopositioner from 310 to 1320 Hz.

figure 24: A Physike Instrumente P-734 nanopositioner was used to evaluate the proposed control strategies. The closed-loop resolution was

approximately 8nm with a range of 0.1mm.

figure 25: Nanopositioner response (in microns) to a 40-Hz triangle wave, with feedforward control (left) and robust control (Right). The new strategy

(right) shows much less performance deterioration when loaded.

B

Page 39: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control38

B.5 DESIGN, ANALYSIS AND CONTROL Of A fAST NANOPOSITIONING STAGE

Project Leader: S.O.R. Moheimani

Researchers: S.S. Aphale, S.O.R. Moheimani and Y.K. Yong

The demand for flexure-based piezoelectric nanopositioning stages is increasing due to their high mechanical bandwidth, large motion range and low cross-coupling between axes. In the field of cell biology, high bandwidth nanopositioning stages are required to monitor biological processes that evolve at a faster pace than current imaging technology. To increase the bandwidth of nanopositioning stages, various approaches have been investigated including improved mechanical design and higher performance control systems.

In this project, a fast XY nanopositioning stage has been designed to have its first dominant mode at 2.5 kHz (see figure 26). Cross-coupling between the two axes is kept to -35 dB, low enough to utilize SISO control strategies for tracking. finite-element-analysis (fEA) is used during the design process to analyse the mechanical resonance frequencies, travel range and cross-coupling between the X and Y axes of the stage. Integral Resonant Control (IRC) is applied in conjunction with a feedforward inversion technique to achieve high-speed and accurate scanning performance. The IRC scheme is a simple yet well-performing technique that adds substantial damping to resonant modes without exciting high frequency dynamics. Together with feedforward control, accurate high-speed scans of up to 400 Hz were achieved.

a)

b)

figure 26: (a) Experimental setup of the XY nanopositioning stage. Capacitive sensors are used to measure the displacement of the stage. (b) fEA simulations of the nanopositioning stage in the X axis. Simulated first resonant mode appears at 2.5 kHz.

B

Page 40: 2008 - University of Newcastle

2008 ANNUAL REPORT 39

B.6 SIMULTANEOUS SENSING AND ACTUATION WITH A PIEZOELECTRIC TUBE SCANNER

Project Leader: S.O.R. Moheimani

Researchers: S.O.R. Moheimani and Y.K. Yong

Piezoelectric tube scanners with quartered external electrodes are the most widely used nanopositioning technology in modern scanning probe microscopes. There has been increasing interest in utilizing feedback control to improve the bandwidth and accuracy of these devices. Non-contact displacement sensors, e.g. capacitive and inductive sensors, have been used for this purpose. However, their measurements contain a significant noise component if operated over large bandwidths. The piezoelectric voltage induced in a tube nanopositioner has been proposed recently as an alternative measure of displacement with a much improved noise figure, up to three orders of magnitude better than capacitive sensors. In this arrangement, an electrode is used to actuate the tube, while the opposite electrode is used as a sensor. This approach has two drawbacks: i) the operating range of the tube is reduced to half, and ii) the tube is not driven symmetrically, thus the opposite sides of the tube experience asymmetric stresses. In this project, a new electrode pattern for piezoelectric tube scanners was developed (see figure 27) which addresses the above problems and allows simultaneous sensing and actuation of the tube in an efficient way.

figure 27: (left) Experimental setup of the new 12-electrode tube scanner. (right) The tube was connection diagram

B

Page 41: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control40

B.7 IDENTIfICATION AND CONTROL Of NEGATIVE IMAGINARY SYSTEMS

Project leader: S.O.R. Moheimani

Researchers: B. Bhikkaji and S.O.R. Moheimani

External Academic Collaborators: I.R. Petersen (NSW, ADfA Campus, Australia)

Many scientific and industrial devices include components that can be classified as flexible structures. flexible structures are susceptible to high amplitude vibrations even in the presence of weak disturbances. These oscillations can result in significant loss of precision and a possible breakdown of the structure if oscillations cross the elastic limit. There is a clear need to damp oscillations that arise in flexible structures.

Under certain conditions, flexible structures with collocated sensor/actuator pairs lead to systems that satisfy the negative imaginary property, i.e., systems which satisfy the following relation,

Here, R(s) denotes the system transfer function.

In this project a System Identification framework was also developed for negative imaginary systems. The identification scheme is based on the subspace method with a negative imaginary constraint imposed.

Integral Resonant Control (IRC) was also developed. This technique provides excellent damping performance and a high degree of robustness for SISO systems with a collocated sensor/actuator pair. A further aspect of this project was to extend IRC control to flexible structures with multiple collocated sensor/actuator pairs (MIMO IRC). It was shown that MIMO IRC also possesses the negative imaginary property and that closed loop stability can be guaranteed.

To validate the proposed identification and control schemes, the developed techniques were successfully applied to an experimental flexible beam. The MIMO IRC control scheme provided excellent attenuation of the beam’s resonance modes. The frequency response of the beam, with and without control, is plotted in figure 28.

figure 28: Magnitude frequency response of the beam from an applied stress to the resulting tip displacement.

i(R (iv) – R* (iv))>_ 0.

B

Page 42: 2008 - University of Newcastle

2008 ANNUAL REPORT 41

B.8 INTEGRAL RESONANT CONTROL Of PIEZOELECTRIC TUBE NANOPOSITIONERS

Project Leader: S.O.R. Moheimani

Researchers: B. Bhikkaji and S.O.R. Moheimani

A vital component of the scanning unit in an Atomic force microscope is a piezoelectric tube nanopositioner. This tube is used for manoeuvring either the probe or sample in a raster pattern. The maximum scanning frequency of a piezoelectric tube scanner is greatly hampered by the presence of resonant modes.

In this project, Integral Resonance Control, developed within the centre, was applied to damp the first resonant modes of a piezoelectric tube. Excellent attenuation of the low-frequency resonance modes was achieved. The corresponding time-domain improvements are demonstrated below in figure 29.

figure 29: Response of a piezoelectric tube to triangular inputs of 10 Hz, 40 Hz and 80 Hz. The

open-loop response is shown on the left, while the response with IRC control is shown on the right.

B

Page 43: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control42

B.9 PRECISE TIP POSITIONING Of fLEXIBLE MANIPULATORS

Project Leader: S.O.R. Moheimani

Researchers: B. Bhikkaji, I.A. Mahmood (Student) and S.O.R. Moheimani

The use of robotic arms to assemble components in the manufacturing industry has increased over the years. Conventional robotic arms have always been structurally rigid and very heavy. However, rigid and heavy robotic arms consume a lot of power, and due to their rigidity, their payload-to-weight ratio is also poor.

The need for high speed manipulation and high payload-to-weight ratio in robotic arms has triggered new research activities related to flexible manipulators (flexible robotic arms). flexible manipulators are light-weight, small-sized and consume a lesser amount of energy for actuation. However, the control system must operate with the presence of lightly damped structural modes.

In this project, a single-link flexible manipulator was fabricated for testing new control schemes. The flexible manipulator is modeled as a single-input two-output system. The system input is torque t while the outputs are the hub angle u and tip deflection. The aim is to move the manipulator tip to a desired location which is the sum of hub angle and beam deflection.

The proposed control system consists of two feedback controllers, a resonant controller and an integral controller. The resonant controller links the hub angle to torque and is designed to damp the flexible modes. The integral controller is used to guarantee zero steady-state tip position error.

The proposed control design was benchmarked against a standard LQG controller. for the same degree of performance, the LQG controller needed more actuation power than the proposed scheme and was sensitive to the presence of unmodeled high-frequency modes. In addition, the LQG controller was also sensitive to the variation of structural resonance frequencies resulting from varying payload mass.

B

figure 30: The experimental flexible manipulator (top) and its closed-loop step response (bottom). The solid line is experimental, which compares well with the simulation (dashed line).

Page 44: 2008 - University of Newcastle

2008 ANNUAL REPORT 43

C. CONTROL SYSTEM DESIGN

Program Goals:

The goal of this program is to develop new techniques for analysing and designing complex control systems. Of special interest are non-linear and non-smooth behaviour, high state dimension and lack of convexity.

C.1 NONLINEAR MODEL PREDICTIVE CONTROL

Project Leader: D.E. Quevedo

Researchers: J.A. De Doná, G.C. Goodwin, J.B. Mare, J. Østergaard and D.E. Quevedo, M.M. Seron and f. Suryawan (Student)

External Academic Collaborator:A. Ahlén (Uppsala Universitet, Sweden)A. feuer (The Technion, Israel) H. Haimovich (Universidad Nacional de Rosario, Argentina)J. Rammage (University of Wollongong, Australia)M. Lazar (Technische Universiteit Eindhoven, The Netherlands)J. Levine (Nationale Supérieure des Mines de Paris, france)

C.1.1 Predictive power control of wireless sensor networks for closed loop control

Researchers: A. Ahlén (Sweden), G.C. Goodwin and D.E. Quevedo

We have studied a networked control architecture where wireless sensors are used to measure and transmit plant outputs to a remote controller. Packet loss probabilities depend upon the time-varying communication channel gains and the transmission powers of the sensors. Within this context, we have developed a centralized stochastic nonlinear model predictive controller. It determines the sensor power levels by trading energy expenditure for expected plant state variance. To further preserve sensor energies, the power controller sends coarsely quantised power increment commands only when necessary. Simulations on measured channel data illustrate the performance achieved by the proposed controller – see Quevedo, Ahlén and Goodwin (2008) in Conference Papers.

C.1.2 A vector quantisation approach to scenario generation for stochastic NMPC

Researchers: A. feuer (Israel), G.C. Goodwin, J. Østergaard and D.E. Quevedo

We have developed a novel technique for scenario generation aimed at closed loop stochastic nonlinear model predictive control. The key ingredient in the algorithm is the use of vector quantisation methods. We have also shown how one can impose a tree structure on the resulting scenarios. We have also described how the scenarios can be used in large scale stochastic nonlinear model predictive control problems. We have tested the idea on a specific problem related to optimal mine planning – see Goodwin, Østergaard, Quevedo and feuer (2008) in Chapters in Books and International Plenary/Keynote Addresses, Goodwin, Seron and Mayne (2008) in Journal Papers.

Maria Seron Program Leader

Graham Goodwin Deputy Program Leader

C

Page 45: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control44

C.1.3 Constrained nonlinear model predictive control

Researchers: J.A. De Doná, H. Haimovich (Argentina), M. Lazar (The Netherlands), J.Levine (france), J.B. Mare, J. Rammage (Wollongong) M.M.Seron and f. Suryawan (Student).

We have investigated a framework for dealing with certain classes of constrained nonlinear model predictive control (MPC) problems which consists in solving a quadratic programming (QP) optimisation at each sampling time. This feature broadens the applicability of nonlinear MPC since many efficient tools for solving QP problems (both numerical algorithms and explicit solutions) are available. One of the contributions has been to show that the optimal control sequence for a nonlinear system of certain features, with a quadratic cost and linear inequality constraints can be computed in exact form via QP provided the optimisation horizon is no larger than a critical quantity that we name the “input-output linear horizon.” In addition, the issue of input to state stability (ISS) with respect to disturbance inputs has been investigated. One proposed solution to guarantee ISS consists of incorporating a set of extra linear inequality constraints into the QP. Thus, the resulting constrained nonlinear MPC scheme has two very attractive properties; namely, simplicity of the solution and guaranteed stability and robustness. The results were published in the Journal Paper, Mare et. al., 2008.

We have also studied the problem of trajectory generation for nonlinear constrained systems. We have developed a novel methodology that combines the differential flatness formalism for trajectory generation of nonlinear systems and the use of an MPC strategy for constraint handling. The methodology consists of a trajectory generator that generates a reference trajectory parameterised by splines, and with the property that it satisfies performance objectives (e.g., satisfies given initial and final conditions, passes through a given set of way-points, etc.). The reference trajectory is generated iteratively in accordance with information received from the MPC formulation. This interplay with MPC guarantees that the trajectory generator receives feedback from present and future constraints for real-time trajectory generation. Thus, the proposed method unites two important properties. firstly, since the trajectories are generated via the flatness parameterisation, with “feedback from the constraints,” they constitute “natural trajectories” for the nominal model to follow. And, secondly, the information generated by an MPC formulation ensures that the system constraints are taken into account. The results have been published in the conference paper De Doná, Suryawan, Seron, and Levine, 2008, and will also appear in a book volume of Springer-Verlag.

C.2 ROBUST MODEL PREDICTIVE CONTROL

Project Leader: M.M. Seron

Researchers: C. Løvaas, M.M. Seron, G.C. Goodwin and S. Hovland

External Academic Collaborators: J.T. Gravdahl (Norwegian University of Science and Technology, Norway) S. Hovland (Student, Norwegian University of Science and Technology, Norway)

C.2.1 Robust output-feedback model predictive control for systems with unstructured uncertainty

Researchers: C. Løvaas, M.M. Seron, G.C. Goodwin, J.T. Gravdahl (Norway) and S. Hovland

We have developed novel results that parameterize a broad class of robust output-feedback model predictive control (MPC) policies for discrete-time systems with constraints and unstructured model uncertainty. The MPC policies we consider employ: (i) a linear state estimator, (ii) a pre-determined feedback gain (iii) a set of “tighter constraints” and (iv) a quadratic cost function in the degrees of freedom and the estimated state. Contained within the class, we find both well-known control policies and policies with novel features. The unifying aspect is that all MPC policies within the class satisfy a robust stability test. The robust stability test is suited to synthesis and incorporates a

C

Page 46: 2008 - University of Newcastle

2008 ANNUAL REPORT 45

novel linear matrix inequality (LMI) condition which involves the parameters of the cost function. The LMI is always feasible under an appropriate small-gain condition on the pre-determined feedback gain and the state estimator. We have shown by means of both theoretical and numerical results, that choosing the cost function parameters subject to the proposed condition often leads to good nominal performance whilst at the same time guaranteeing robust stability.

We have also developed novel results linking model predictive control (MPC) and minimax optimal control theory. Specifically, we have shown that the closed-loop optimal solutions of a particular class of minimax optimal control problems are a class of typical MPC policies, for linear discrete-time systems with constraints and disturbance inputs. We have also developed conditions which ensure that the inverse optimal MPC policies achieve a prescribed (regional) l 2 gain from the disturbance input.

We have also developed a robust output-feedback model predictive control (MPC) design for a class of open-loop stable systems with hard input- and soft state constraints. The proposed output-feedback design is based on a linear state estimator and a novel parameterization of the soft state constraints that has the advantage of leading to optimisation problems of prescribable size. Robustness against unstructured model uncertainty is obtained by choosing the cost function parameters so as to satisfy a linear matrix inequality condition. We have also shown, by means of both theoretical and numerical results, that the proposed design methodology often leads to good nominal performance.

We have applied the above idea to develop a systematic procedure for obtaining closed-loop stable output-feedback model predictive control based on reduced-order models. The design uses linear state estimators, and applies to open-loop stable systems with hard input- and soft state constraints. Robustness against the model reduction error is obtained by choosing the cost function parameters so as to satisfy a linear matrix inequality condition. We have also shown by means of an example, that performance is maintained even when the model reduction error is relatively large – see Løvaas, Seron and Goodwin (2008) in Journal Papers, Journal Papers (accepted for publication) and Conference Papers, Hovland, Løvaas, Gravdahl and Goodwin (2008) in Conference Papers.

C.3 VIRTUAL LABORATORIES fOR CONTROL EDUCATION

Project Leader: G.C. Goodwin

Researchers: G.C. Goodwin, J.S. Welsh, A.M. Medioli and f. Sobora

We have further developed our virtual laboratory for control education. We now have available the following laboratories:

A. Virtual Industrial Laboratories

1. Continuous Caster Package: This contains:

(a) Continuous Caster – Classical Control Design

(b) Continuous Caster – Nonlinear Issues

2. Rolling Mill Package: This contains:

(a) Rolling Mill – System Modelling and Classical Control

(b) Rolling Mill – Soft Sensors

(c) Rolling Mill – Periodic Disturbances and Observer Design

(d) Rolling Mill – Kalman filteringVirtual laboratories stand at the 17th IfAC World

Congress held in Seoul, South Korea 6-11 July 2008.

C

Page 47: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control46

3. Rocket Dynamics Package: This contains:

(a) Rocket Dynamics

(b) Rocket Controller Design

4. Paper Machine Package: This contains:

(a) Cross Directional Control of Web forming Processes – Interaction and Simple PID Control

(b) Cross Directional Control of Web forming Processes – Actuator Saturation

(c) Cross Directional Control of Web forming Processes – Robustness

B. Virtual Benchtop Laboratories

5. Audio Signal Processing Package: This contains:

(a) Audio Signal Processing – Optimal Noise Shaping Quantiser

(b) Audio Signal Processing – Bode Sensitivity Integrals

6. Electromechanical Servomechanism Package: This contains:

(a) Electromechanical Servomechanism

(b) Resonant Electromechanical Servomechanism

7. Fluid Tanks Package:

This contains:

(a) fluid Tanks – Single Tank Control

(b) fluid Tanks – Coupled Tanks Control

further details are available at: http://www.virtual-laboratories.com

See Goodwin, Sobora, Bastiani (2008) in Books, Welsh, Daredia, Sobora, Vlacic and Goodwin (2008) in Conference Papers.

C.4 ACTUATOR fAULT TOLERANT CONTROL

Project Leader: M.M. Seron

Researchers: J.A. De Doná, M.M. Seron, C. Ocampo-Martinez and A. Yetendje (Student)

External Academic Collaborators: J.-J. Martinez (Laboratoire d’Automatique de Grenoble, france) S. Olaru (Laboratoire d’Automatique de Grenoble, france) M. Romero (Universidad Nacional de Rosario, Argentina)

This project is part of a general project on fault tolerant control, co-funded by the ARC discovery grant DP0881419: fault tolerant multisensor feedback control. The latter project mainly deals with the multisensor fault tolerant control problem whereas here we focus on strategies to treat actuator faults.

We have developed a new actuator fault tolerant control strategy based on invariant set computation. The proposed scheme employs a bank of observers which match the different fault situations that can occur in the plant. Each of these observers produces an estimation error with a distinctive behaviour when the observer matches the current fault situation in the plant. With the information of the estimation errors from each of the considered observers, a fault diagnosis and isolation (fDI) module is able to reconfigure

C

Page 48: 2008 - University of Newcastle

2008 ANNUAL REPORT 47

the control loop by selecting the appropriate stabilising controller from a bank of precomputed control laws, each of them related to one of the considered fault models. The decision criteria of the fDI is based on the computation of invariant sets of the estimation errors for each fault scenario and for each control configuration. We have derived conditions for the design of the fDI module and for fault tolerant closed-loop stability. These results were published in the conference paper Ocampo-Martinez, De Dona and Seron, 2008. In addition, one journal and one conference papers addressing refinements and extensions of the strategy have been submitted for publication.

C.5 LIMITATIONS IN fEEDBACK CONTROL OVER COMMUNICATION CHANNELS

Project Leader: J.H. Braslavsky

Researchers: J.H. Braslavsky, A.J. Rojas and R.H. Middleton

External Academic Collaborators: J.S. freudenberg (University of Michigan, USA) J.I. Yuz (Universidad Técnica federico Santa María, Chile)

C.5.1 Infimal feedback capacity for a class of additive coloured gaussian noise channels

Researchers: J.S. freudenberg (USA), R.H. Middleton and A.J. Rojas

This work has studied the infimal signal-to-noise ratio (SNR) required for stabilisation of a linear time invariant (LTI) scalar unstable plant over a class of additive coloured Gaussian noise channels. We have applied recent results in the literature to obtain the feedback capacity of such a class of channels. We have proven that the infimal SNR constrained LTI solution, when dealing with a scalar unstable plant, achieves a channel feedback capacity equal to the infimal rate of transmission required for stability. The optimality of such channel feedback capacity is a non trivial result since we consider additive 1st order moving average (MA) and autoregressive moving average (ARMA) coloured noise.

C.5.2 Repeated poles in feedback over a class of signal-to-noise ratio constrained channels

Researchers: A.J. Rojas and J.I. Yuz (Chile)

This work studies the H2 norm in feedback systems subject to signal-to-noise ratio constraints. We have obtained a closed form expression for the squared norm in the orthogonal complement of the space H2 of a partial fraction expansion with repeated unstable poles. We have also obtained a closed form expression for the squared H2 norm of a partial fraction expansion with repeated stable poles. As an application we use the result on norm on the orthogonal complement to H2 to extend the closed form solution of the discrete time linear time invariant (LTI) signal-to-noise ratio (SNR) constrained problem to the case of repeated unstable poles in the plant model.

C

Page 49: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control48

figure 31: Minimum variance control over a Gaussian channel: Estimation over a channel with feedback.

figure 32: Disturbance rejection in channel SNR constrained feedback control: Output feedback control over an additive coloured Gaussian noise channel with memory.

C.5.4 Disturbance rejection in channel signal-to-noise ratio constrained feeback control

Researchers: J.H. Braslavsky, J.S. freudenberg (USA), R.H. Middleton and A.J. Rojas

Communication channels impose a number of obstacles to feedback control. One recent line of work considers the problem of feedback stabilisation subject to a constraint on the channel signal-to-noise ratio (SNR). It has been shown for continuous-time systems that the optimal control problem of achieving the infimal SNR can be formulated as a linear quadratic Gaussian (LQG) control problem with weights chosen as in the loop transfer recovery (LTR) technique. The present paper extends this formulation to: discrete- time systems; communications over channels with memory; and input disturbance rejection. By using this formulation, we derive exact expressions for the linear time invariant (LTI) controller that achieves the infimal SNR under the effect of time-delay and additive coloured noise. We then quantify the infimal SNR required for both stabilisation and input disturbance rejection for a relative degree one, minimum phase plant and a memoryless Gaussian channel.

CC.5.3 Minimum variance control over a Gaussian communication channel

Researchers: J.H. Braslavsky, J.S. freudenberg (USA) and R.H. Middleton

This work focuses on the problem of minimizing the response of a plant output to a stochastic disturbance using a control law that relies on the output of a noisy communication channel. We analyse a lower bound on the performance achievable at a specified terminal time using nonlinear time-varying communication and control strategies, and show that this bound may be achieved using strategies that are linear.

The communication and control strategies derived are optimal with respect to minimizing the mean square plant output and the variance of the output estimation error. However, in contrast to the standard LQG problem, we show that the tasks of control and estimation cannot be separated. We have also obtained communication and control strategies that are valid over an infinite horizon. These will be suboptimal for any specific time, but may yield better transient properties.

Page 50: 2008 - University of Newcastle

2008 ANNUAL REPORT 49

10−2 10−1 100 101−40

−20

0

20

ω (rad/sec)

|H(jω

)| dB 1st Vehicle

20th Vehicle

0 50 100 150 200−2

−1

0

1

2

Time (sec)

e(t)

1st Vehicle

20th Vehiclefigure 33: String stability under bidirectional

(preceding and following vehicles) communication control. String sensitivity peak (top) is reduced with the vehicle position, at the expense of larger, longer

transient performance (bottom).

C.5.5 Performance limitations in distributed control systems

Researchers: J.H. Braslavsky and R.H. Middleton

This work focuses on the derivation of sufficient conditions for string instability in a string of linear time-invariant autonomous vehicles with constrained communication structures.

String instability is a problem arising in diverse applications, including intelligent vehicle highway systems, irrigation channel systems, and supply chain management. The problem is manifested as an amplification of disturbances along the string that become unbounded as the length of the string grows.

Our approach is based on Bode integral constraints, extending earlier works to the case of heterogeneous, non-nearest neighbour and semi-rigid formation situations. The vehicles are assumed to be controlled autonomously and are subject to a rigid or semi-rigid formation policy. The individual controllers are assumed to have a limited range of forward and backward communication with other vehicles.

Sufficient conditions are obtained that imply a lower bound on the maximum peak of the frequency response magnitude of the transfer function which maps a disturbance to the leading vehicle to a vehicle in the chain. This lower bound quantifies the effect of:

n spacing separation policy,n intervehicle communication policy, andn vehicle settling response performance,

The results are useful to characterise tradeoffs and infeasible specifications in control and policy design. The analysis also points to factors that are beneficial to string stability, such as:

n increasing communication ranges, specially forward communication,n increasing loop high frequency response, andn relaxing transient (low frequency) performance requirements.

C

Page 51: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control50

C.6 CONTROL Of POWER PLANTS

Project Leader: G.C. Goodwin

Researchers: E.I. Silva (Student) and G.C. Goodwin

External Academic Collaborators:V. Wertz (Université Catholique de Louvain, Belgium)B. Codrons (Université Catholique de Louvain, Belgium)

C.6.1 Performance limitations arising in the control of power plants

Researchers: B. Codrons (Belgium), V. Wertz (Belgium), G.C. Goodwin and E.I. Silva

We have developed results on performance limitations for direct fixed coal power plants. A specific feature of this system is the existence of a very large input delay between one of the inputs, namely coal flow, and the two outputs, load and vapour pressure. This problem motivates the examination of tracking performance limitations in one process variable when another process variable is constrained. Our main result makes explicit the performance trade-off between the two conflicting objectives, and also links the achievable performance to the delay structure of the plant. These results give insights into the benefits of MIMO control for power plants and into the necessary trade-off between fast tracking of load step changes and the need for minimizing the variations of the vapour pressure around its nominal value. The results provide a benchmark against which practical controller designs for power plants can be assessed – see Wertz, Silva, Goodwin and Codrons (2008) in Conference Papers.

CC.7 ULTIMATE BOUNDS IN PERTURBED SYSTEMS

Researchers: M.M. Seron

External Academic Collaborators:E. Kofman (University of Rosario, Argentina)H. Haimovich (University of Rosario, Argentina)f. fontenla (University of Rosario, Argentina)

In previous reports we presented a new systematic method to compute ultimate bounds for perturbed systems. The method is based on a componentwise analysis of the system in modal coordinates and thus exploits the system geometry as well as the perturbation structure without requiring calculation of a Lyapunov function for the system. We have recently made an in-depth analysis and provided improvements of the method, aiming at reducing the conservatism of the procedure even further, hence leading to tighter bounds. These improvements were presented in the Conference Paper Haimovich, Kofman and Seron, 2008.

We have also combined the above method with a technique for eigenvalue/eigenvector assignment by state feedback in a new control design algorithm for perturbed multiple-input systems. The new algorithm is systematic in nature and guarantees any desired componentwise ultimate bound on the state for systems with various types of uncertainties, including uncertain time-delays in the feedback loop. The results have been published in the Journal Paper Kofman, Seron and Haimovich, 2008.

In addition, we have extended the results to a class of perturbed feedback linearisable nonlinear systems with matched perturbations. for these systems, we have developed a systematic design procedure to compute a state feedback control that ensures a prescribed ultimate bound for the closed-loop system states. The procedure combines nonlinear state feedback linearising control with a state-feedback matrix computed via the eigenstructure assignment method mentioned in the paragraph above. The results were presented in the Conference Paper Kofman et.al., 2008.

C.8 QUANTISED CONTROL

Researchers: M.M. Seron

External Academic Collaborators: H. Haimovich (University of Rosario, Argentina).

This project deals with quadratic stabilisation of discrete-time systems with quantisers. In previous years we have obtained an explicit geometric characterisation of quadratically stabilising state feedback laws that are based on the use of multivariable quantisers of

minimum dimension. Using the geometric characterisation, we have developed a systematic design procedure for finite-density quadratically-stabilising quantisers for multiple-input (MI) systems. The resulting quantisers have a simple geometric structure and can be implemented via simple function evaluations. The results were published in the Journal Paper Haimovich and Seron, 2008.

Page 52: 2008 - University of Newcastle

2008 ANNUAL REPORT 51

D. SIGNAL PROCESSING

Program Goals:

This program focuses on model-based signal processing. Research problems include physical modelling, system identification, model validation, prediction, filtering, and signal recovery. In the past, we have studied various signal processing problems in adaptive control, Kalman filtering, communications channel equalization, and multi-user detection for wireless communications. In 2008, we have focused our research in the following areas: system identification, quantisation, dual-stage control, time-frequency analysis and communication systems. Specific applications include commodity price modelling, high-precision positioning and high-speed analog-top-digital converter design.

D.1 SYSTEM IDENTIfICATION

Project Leader: J.-C. Agüero

Researchers: G.J. Adams, J.-C. Agüero, G.C. Goodwin, C. Rojas (Student) and J.S. Welsh

External Academic Collaborators: M. Alamir (Institut National Polytechnique de Grenoble, france), P.M.J. Van den Hof (Delft University of Technology, The Netherlands) J.I. Yuz (Universidad Técnica federico Santa María,Chile)

D.1.1 Experiment design for system identification

This research is partly funded under an ARC Discovery Grant.

We have a large research program aimed at gaining a better understanding of experiment design. There are several sub-projects within this research program.

One of the topics studied in 2008 is the equivalence between least costly and traditional experiment design for control. We consider experiment design problems for both open and closed loop systems. In open loop, equivalence is established for three specific cases, relating to different parameterisations of the covariance expression (i.e. finite and high order approximations) and model structure (i.e. dependent and independently parameterised plant and noise models). In the closed loop setting, we consider only finite order covariance expressions. H∞ performance specifications for control are used to determine the bounds on the covariance expression for both the open and closed loop cases – see Rojas, Agüero, Welsh and Goodwin (2008) in Journal Papers.

On the other hand, we also have developed fundamental integral limitations on the variance of estimated parametric models, for both open and closed loop identification. As an application of these results we have shown that, for multisine inputs, a well known asymptotic (in model order) variance expression provides upper bounds on the actual variance of the estimated models for finite model orders. The fundamental limitations established here give rise to a ‘water-bed’ effect – see Rojas, Agüero, Welsh (2009) in Journal Papers.

D.1.2 On parameter estimation of the Schwartz-Smith short-term/long-term model

Researchers: M. fu and Xin Tai (Visiting Student)

The short-term/long-term model proposed by Schwartz and Smith in 2000 is widely used in modelling commodity prices. A key and nontrivial problem in this modelling technique is how to estimate the model parameters. This work considers the parameter estimation problem based on the maximum likelihood criterion and proposes a method to simplify the task. Two components are contained in the proposed method: one to do with re-parameterization and one to do with separating the parameter set so that one part can be solved directly using least-squares and another part using nonlinear optimization. The effectiveness of the proposed method has been demonstrated via numerical tests.

Minyue fu Program Leader

Juan-Carlos Agüero Deputy Program Leader

D

Page 53: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control52

D.1.3 Identifiability of errors in variables dynamic systems

Researchers: J.-C. Agüero and G.C. Goodwin

There has been substantial research carried out on the errors in variables (EIV) identifiability problem for dynamic systems. These results are spread across a significant volume of literature. We have developed a single theorem which compactly summarizes many of the known results. The theorem also covers several cases which we believe to be novel. We have analysed single input single output systems using second order properties. We have also extended the results to a class of multivariable systems.

We have also developed novel results related to the identifiability of EIV dynamic systems based on exploiting properties of non-stationary data. We have analysed single-input single-output systems using second order properties. Our results show that, it is possible to establish identifiability of EIV systems under mild conditions when the data is non-stationary – see Agüero and Goodwin (2008) in Journal Papers and Conference Papers.

D.1.4 Robust identification of process models from plant data

Researchers: G.J. Adams, J.-C. Agüero, G.C. Goodwin, C.R. Rojas, J.S. Welsh and J.I. Yuz (Chile)

A precursor to any advanced control solution is the step of obtaining an accurate model of the process. Suitable models can be obtained from phenomenological reasoning, analysis of plant data or a combination of both. We have a large research program aimed at the problem of estimating (or calibrating) models from plant data. A key goal is to achieve robust identification. By robust we mean that small errors in the hypotheses should lead to small errors in the estimated models. We argue that, in some circumstances, it is essential that special precautions, including discarding some part of the data, be taken to ensure that robustness is preserved. We have illustrated these ideas using practical case studies – see Yuz and Goodwin (2008) in Chapters in Books, Goodwin, Agüero, Welsh, Yuz, Adams and Rojas (2008) in Journal Papers.

D.1.5 Virtual closed loop identification: A generalized tool for identification in closed loop

Researchers: J.-C. Agüero, G.C. Goodwin and P.M.J. Van den Hof (The Netherlands)

We have proposed a virtual closed loop model parameterization to perform system identification. This parameterization is designed to achieve specific goals. We show that the method includes, as special cases, known methods for closed loop identification and also offers additional flexibility. We have analysed the ramifications of the new tailor-made parameterization for systems operating in closed loop. The approach exploits a property of Box-Jenkins models in order to minimize the bias arising from feedback and noise model mismatch – see Agüero, Goodwin and Van den Hof (2008) in Conference Papers.

D.1.6 Relative error issues in sampled data models

Researchers: J.-C. Agüero, G.C. Goodwin and J.I. Yuz (Chile)

Most real world systems operate in continuous time. However, to store, analyse or transmit data from such systems the signals must first be sampled. Consequently there has been on-going interest in sampled data models for continuous time systems. The emphasis in the literature to-date has been on three main issues namely the impact of folding, sampled zero dynamics and the associated model error quantification. Existing error analyses have almost exclusively focused on unnormalised performance. However, in many applications relative errors are more important. for example, high performance controllers tend to invert the system dynamics and consequently relative errors underpin closed loop performance issues including robustness and stability. This has motivated us to examine the relative errors associated with several common sampled data model types. This analysis reveals that the inclusion of appropriate zero dynamics is essential to ensure that the relative error converges to zero as the sampling period is reduced – see Goodwin, Yuz and Agüero (2008) in Conference Papers.

D.1.7 Redundancy vs multiple starting points in nonlinear systems related inverse problems

Researchers: M. Alamir (france), G.C. Goodwin and J.S. Welsh

Nonlinear inverse problems arising in dynamic system state/parameter estimation are generally nonconvex and possess multiple local minima that may threaten the convergence of global optimisation routines. This problem is generally addressed by a multiple starting point based algorithm. An alternative approach is to use the recently proposed concept of safe redundancy in order to derive an algorithm that crosses singularities by exploiting the particular nature of dynamic inverse problems. In fact, it is possible to combine the two algorithms such that the properties of each are utilised in a hybrid algorithm. The underlying idea in safe redundancy is not to choose a good cost function but to have different cost functions, that share the same global minimum but have no reason to share all the local minimum, and make them cooperate to reach the desired solution – see Alamir, Welsh and Goodwin in Journal Publications.

D.1.8 On useful redundancy in robust optimal experiment design for nonlinear system identification

Researchers: M. Alamir (france), G.C. Goodwin and J.S. Welsh

This research is partly funded under an ARC Discovery Grant.

In particular, a recently introduced redundancy property associated to inverse problems, has been for dynamic systems is heavily exploited to guarantee global convergence. The project considers general discrete-time nonlinear systems in which measurements are affected by bounded noise – see Alamir, Welsh and Goodwin in Conference Publications.

D

Page 54: 2008 - University of Newcastle

2008 ANNUAL REPORT 53

D.2 DUAL-STAGE SYSTEMS

Project Leader: Minyue fu

Researchers: M. fu, A. Salton (Student) and J. Zheng

External Academic Collaborators: Y. Wang (Nanyang Technological University, Singapore) C. Du (Agency for Science, Technology and Research (A*STAR) Data Storage Institute (DSI), Singapore) Y. Guo (Nanyang Technological University, Singapore) L. Xie (Nanyang Technological University, Singapore)

D.2.1 Development of a linear motion dual-stage actuator

Researchers: J. Zheng and M. fu

A dual-stage actuator (DSA) servo system is characterized by a structural design with two actuators connected in series along a common axis. The primary actuator (coarse actuator) is of long travel range but with poor accuracy and slow response time. The secondary actuator (fine actuator) is typically of higher precision and faster response but with a limited travel range. By combining the DSA system with properly designed servo controllers, the defects of one actuator are compensated by the merits of the other, resulting in an improved performance. We have designed a linear motion DSA where the primary stage is driven by a linear motor (LM) and the secondary stage is driven by a piezoelectric actuator (PZT). Experimental results demonstrate a significant improvement both in settling time reduction and disturbance rejection when comparing the proposed design with a traditional single-stage actuator system.

D.2.2 Development of a rotational motion dual-stage actuator

Researchers: J. Zheng, A. Salton and M. fu

While the use of dual-stage actuators (DSAs) is popular in hard disk drives (HDDs), some other applications that make use of this structural design include XY positioning tables, macro/micro robot manipulators, machine tools, wafers alignment in microlithography, among others. Motivated by the fact that several other applications may benefit from long range over a compact space (such as HDDs), we have designed and manufactured the rotational motion dual-stage actuator as depicted in figure 34. This system is based in the same principle as the linear motion DSA, where the primary stage, driven by a voice coil motor (VCM), provides the system with a large range, and the secondary stage, driven by a piezoelectric actuator (PZT), achieves a higher precision and faster response. The mechanical design of the rotational DSA, however, is different and more sophisticated than the one of the linear motion DSA. In order to achieve a rotational motion the secondary stage must bend the structure of the actuator, which requires the system to have complex mechanical structure. To overcome this difficulty we have made use of a finite element method software tool to aid our mechanical design. Simulation and experimental results demonstrate the effectiveness of the proposed DSA system.

figure 34: The rotational motion dual stage actuator.

D.2.3 Nonlinear tracking control for a hard disk drive

Researchers: M. fu, C. Du (Singapore), Y. Wang (Singapore) and J. Zheng

This work studies a nonlinear tracking control method for a hard disk drive dual-stage actuator (DSA) system that consists of a voice coil motor (VCM) actuator and a piezoelectric (PZT) microactuator. Conventional track seeking controllers for DSA systems were generally designed to enable the VCM actuator to approach the target track without overshoot. However, we observe that this strategy is unable to achieve the minimal settling time when the target tracks are beyond the PZT actuator stroke limit. To further reduce the settling time, we design the VCM actuator controller to yield a closed-loop system with a small damping ratio for a fast rise time and certain allowable overshoot. Then, a composite nonlinear control law is designed for the PZT actuator to reduce the overshoot caused by the VCM actuator as the system output approaches the target track. Experimental results show that the proposed dual-stage servo outperforms the conventional dual-stage servo in short-span seeking and, additionally, achieves better track following accuracy than the VCM only single-stage servo.

figure 35: Illustration of the proposed control strategy.

D

Page 55: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control54

D.2.4 Development of an extended reset controller and its experimental demonstration

Researchers: M. fu, Y. Guo (Singapore), Y. Wang (Singapore), L. Xie (Singapore) and J. Zheng

Reset control aims at enhanced performance that cannot be obtained by linear controllers. The conventional reset control is simple for implementation by resetting some of its controller states to zero when its input meets a threshold. However, it is found that in some cases the enhanced performance of conventional reset control is still limited such as with only partial reduction of the overshoot in a step reference response. Thus, the stability analysis and design of the reset control system are extended, where the reset time instances are prespecified and the controller states are reset to certain non-zero values, which are calculated online in terms of the system states for optimal performance. Experimental results on a piezoelectric positioning stage demonstrate that the extended reset control can further reduce the overshoot and thus achieve shorter settling time than the conventional reset control. Robustness tests against various step levels, disturbance and sensor noise have also been developed.

D.2.5 A reset state estimator using an accelerometer for enhanced motion control with sensor quantisation

Researchers: M. fu and J. Zheng

Sensor quantisation is a key factor that deteriorates the tracking performance of positioning systems with low-resolution optical encoders. This work presents a method to improve the performance of such systems by merging an accelerometer of low cost. firstly, to reject the external disturbance, friction force and system perturbations, we design a disturbance observer (DOB) based on acceleration signals. Secondly, a reset kinematic state estimator (RKSE) is designed using acceleration signals to make the state estimate immune to both system perturbations and input disturbances. Thirdly, a state feedback controller is designed based on the internal model principle (IMP) for accurate sinusoidal reference tracking.

Simulations and experimental results are used to demonstrate the effectiveness of the proposed control method for tracking position reference commands and its robustness to system uncertainties.

D.3 QUANTISATION

Project Leader: D.E. Quevedo

Researchers: G.C. Goodwin, S. Derpich (Student), D. Marelli, J. Østergaard, D.E. Quevedo and E.I. Silva (Student)

External Academic Collaborators: H. Bölcskei (ETH Zurich, Switzerland)

D.3.1 Feedback quantisers

Researchers: S. Derpich, J. Østergaard, G.C. Goodwin, D.E. Quevedo, E.I. Silva

We have developed novel results on perfect reconstruction feedback quantisers (PRfQs), i.e., noise-shaping, predictive and sigma-delta A/D converters whose signal transfer function is unity. Our analysis of this class of converters is based upon an additive white noise model of quantisation errors. Our key result is a formula that relates the minimum achievable MSE of such converters to the signal-to-noise ratio (SNR) of the scalar quantiser embedded in the feedback loop. This result allows us to obtain analytical expressions that characterize the corresponding optimal filters. We also show that, for a fixed SNR of the scalar quantiser, the end-to-end MSE of an optimal PRfQ which uses the optimal filters (which for this case turn out to be IIR) decreases exponentially with increasing oversampling ratio. Key departures from earlier work include the fact that fed back quantisation noise is explicitly taken into account and that the order of the converter filters is not a-priori restricted.

We have also developed results on scalar feedback quantisation (SQf) with uniform quantisers. We focus on general SfQ configurations where reconstruction is via a linear combination of frame vectors. Using a deterministic approach, we derive two necessary and sufficient conditions for SfQ to be optimal, i.e., to produce, for every input, a quantised sequence that is

a global minimiser of the 2-norm of the reconstruction error. The first optimality condition is related to the design of the feedback quantiser, and can always be achieved. The second condition depends only on the reconstruction vectors, and is given explicitly in terms of the Gram matrix of the reconstruction frame. As a by-product, we also show that the first condition alone characterizes scalar feedback quantisers that yield the smallest MSE, when one models quantisation noise as uncorrelated, identically distributed random variables.

We have also characterized the rate-distortion function for zero-mean stationary Gaussian sources under the MSE fidelity criterion and subject to the additional constraint that the distortion is uncorrelated to the input. The solution is given by two equations coupled through a single scalar parameter. This has a structure similar to the well known water-filling solution obtained without the uncorrelated distortion restriction. Our results fully characterize the unique statistics of the optimal distortion. We also show that, for all positive distortions, the minimum achievable rate subject to the uncorrelation constraint is strictly larger than that given by the un-constrained rate-distortion function. This gap increases with the distortion and tends to infinity an zero, respectively, as the distortion tends to zero and infinity – see Derpich, Silva, Quevedo and Goodwin (2008) in Journal Papers, Derpich, Quevedo and Goodwin (2008) in Conference Papers, Silva, Quevedo and Goodwin (2008) in Conference Papers.

D.3.2 Quantisation of filter bank frame expansions through moving horizon optimisation

Researchers: H. Bölcskei (Switzerland), G.C. Goodwin and D.E. Quevedo

We have developed a novel approach to quantisation in oversampled filter banks. The new technique is based on moving horizon optimisation, does not rely on an additive white noise quantisation model and allows one to explicitly enforce stability of the associated nonlinear feedback loop. Moreover, the quantisation structure that we have proposed includes sigma-delta and linear predictive subband quantisers as a special case and, in general, outperforms them – see Quevedo, Bolcskei and Goodwin (2008) in Journal Papers.

D

Page 56: 2008 - University of Newcastle

2008 ANNUAL REPORT 55

MoDEl APPRoxiMATion ERRoR CoMPlExiTy [mult./sample]

fIR 40.63e-3 135

PZ 15.29e-3 81

SB 16.55e-3 37

0 531 1255 2711 6414 26646−40

−30

−20

−10

0

10

20

Frequency [Hz]

Ampl

itde

[dB]

True HRTFFIR modelPole−zero modelSubband modelfigure 36: Head-related transfer function (HRTf)

implementation using a finite impulse response (fIR) model, a pole-zero model (PZ) and a subband

(SB) model. Left: frequency response. Right: Approximation error and complexity.

D.4 TIME-fREQUENCY ANALYSIS

Project Leader: M. fu

Researchers: M. fu, K. Mahata and D.E. Marelli

External Academic Collaborators: M. Aramaki (CNRS-Institut de Neurosciences Cognitives de la Méditerranée, Marseille, france) P. Balazs (Austrian Academy of Sciences, Austria) R. Kronland-Martinet (CNRS-Laboratoire de Mécanique et d’Acoustique, Marseille, france. P. Majdak (Austrian Academy of Sciences, Austria) I. Raeburn (University of Wollongong, Australia) C. Verron (Orange Labs, Lannion, france.

D.4.1 Linear system modelling in the subband domain

Researchers: P. Balazs (Austria), M. fu, P. Majdak (Austria) and D.E. Marelli

The subband technique permits a numerically efficient implementation of a linear system. This is done by putting in cascade an analysis filterbank, followed by a transfer matrix (the subband model) and a synthesis filterbank. To exploit the potential of this technique we designed the subband model, the analysis and the synthesis filterbanks, to minimize a statistical least-mean squares criterion; and we included in the design procedure, a sparsity criterion on the subband model. By doing so, we minimize the number of non-zero coefficients of the subband model, and consequently its implementation complexity. In figure 36 we show the performance of the proposed subband design when implementing head-related transfer functions (HRTf). The proposed approach yields significant computational savings in comparison with pole-zero models, or finite impulse response models implemented using fast convolution algorithms.

D

Page 57: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control56

D.4.2 High-speed analog-to-digital converter design

Researchers: M. fu, K. Mahata and D.E. Marelli

High speed analog-to-digital converters (ADCs) can be realized using the so-called multi-channel architecture, which combines the outputs of a number of slow speed ADCs. A research challenge is how to combine the output from different channels to reconstruct the desired samples. This design is typically done by assuming that the input signal is bandlimited, leading to a computaionally expensive design. We proposed an alternative approach by dropping the bandlimited assumption and doing the design in a statistically optimal sense. We have applied this approach in both, time-interleaved ADCs, and for hybrid-filterbank ADCs. The proposed designs lead to superior performances in terms of complexity when compared with other available approaches.

D.4.3 Efficient sound synthesis for real-time applications

Researchers: M. Aramaki (france), R. Kronland-Martinet (france), D.E. Marelli and C. Verron (france)

The inverse fast fourier transform (IffT) synthesis is a flexible and efficient method for synthesizing digital sound. It consists in approximating the spectral shape of the desired sound in a sequence of overlapping blocks of samples. The application of the IffT method is limited by its inherent time/frequency resolution tradeoff, since its block size needs to be longer than the noise auto-correlation function, which is often not compatible with the generation of short transient signals. We have shown that the IffT method can be re-casted as the implementation of a time-varying filter in the subband domain. Using this, we were able to overcome the aforementioned limitation by designing an appropriate subband model for the time-varying filter.

D.4.4 Proper group actions in abstract harmonic analysis

Researcher: D.E. Marelli, I. Raeburn (Wollongong)

Abstract harmonic analysis generalizes the study of continuous-, discrete-time or multidimensional functions, to functions with domain on a topological group (space), which take values (have range) on a C-star algebra rather than on the real or complex numbers. A research line within this context studies the properties of these functions when a group action (a set of automorphisms with a group structure) is applied on their range. A famous theorem of Green asserts that in the case of commutative C-star algebras, if a proper and free action acts on the maximal ideal space, then the crossed product between the acting group and the C-star algebra is Morita equivalent to the functions on the orbit space. Rieffel used this result to abstract the concepts of properness and freeness (which he called saturation) to group actions acting on general (non-commutative) C-star algebras. When these concepts are applied to a particular case of non-commutative C-star algebras called graph algebras, they exhibit a striking parallel with the commutative case. We studied the properties of proper actions which are not saturated. In the commutative case, we showed that not the whole crossed product, but a proper ideal of it satisfies the above Morita equivalence, if and only if the action is not free. Also, we identified this ideal by describing the primitive ideals which contain it. In the non-commutative case, we showed that while the ideal can certainly be proper, there are situations where the action is not free but the ideal is the whole crossed product. Moreover, even in the later case, Morita equivalence may not hold. Our results suggest that the aforementioned parallel completely breaks down in the absence of saturation.

D

Page 58: 2008 - University of Newcastle

2008 ANNUAL REPORT 57

D.5 COMMUNICATION SYSTEMS

Project Leader: G.C. Goodwin

Researchers: J.A. De Doná, M. fu, G.C. Goodwin, J. Ning, D.E. Quevedo and M. Wang,

External Academic Collaborators: B.S. Krongold (The University of Melbourne, Australia)

D.5.1 A complex-baseband active-set approach for tone reservation PAR reduction in OFDM systems

Researchers: G.C. Goodwin, B.S. Krongold (Australia), D.E. Quevedo and M. Wang

We have developed an active-set approach for PAR reduction via time reservation in complex-baseband orthogonal frequency division multiplexing (OfDM) systems. In the complex-baseband model, the optimisation problem of tone reservation turns out to be a quadratically constrained quadratic program (QCQP), which is computationally prohibitive for practical implementations. To address this problem, we consider the complex-valued samples of an OfDM symbol as sectors in the complex plane and develop an iterative clipping algorithm. In our approach, PAR reduction is achieved by minimizing the radius of the circle that enclosed all samples, thus minimizing the maximum power peak. Simulation results show that the proposed algorithm can achieve near-optimal performance with fast convergence and limited complexity – see Wang, Quevedo, Goodwin and Krongold (2008) in Conference Papers.

D.5.2 A multi-step detector for linear ISI-channels incorporating de grees of belief in past estimates

Researchers: J.A. De Doná, G.C.Goodwin and D.E. Quevedo

We have formulated the channel equalization problem in the framework of constrained maximum likelihood estimation. This allows us to highlight key issues including the need to summarize past data

and to apply a finite alphabet constraint over a sliding optimisation window. The approach leads to embellishments of the usual (non-adaptive) decision feedback equalizer and its multi-step extensions. It includes a provision for degrees of belief in past estimates, which addresses the problem of error propagation – see Quevedo, Goodwin and De Dona (2008) in Journal Papers.

D.5.3 EM-based receiver design for uplink MIMO-OFDMA systems

Researchers: G.C. Goodwin, D.E. Quevedo and M.Wang

We have proposed an iterative receiver for uplink multiple-input multiple-output (MIMO) orthogonal frequency division multiple access (OfDMA) systems based on the expectation maximization (EM) algorithm. Iterating between the E-step and the M-step, the EM-based receiver updates the channel estimates and, refines data detection by increasing the likelihood function. Practical implementation issues have also been considered: space-time block-coding (STBC) is incorporated to improve system performance against fading; a reduced-complexity algorithm is proposed, which simplifies the computation whilst not compromising performance. Simulation results show that the performance of the proposed receiver can approach coherent detection with affordable computational cost – see Wang, Goodwin and Quevedo (2008) in Conference Papers.

D.5.4 A new method for lowering the error floors of non-binary LDPC codes

Researchers: M. fu and J. Ning

Non-binary low-density parity-check (LDPC) codes are good candidates for forward-error-control code with short block lengths (normally less than 5,000 bits) and can substantially outperform their binary counterparts. In this work, we propose a new method for improving the performance of non-binary LDPC codes of short block lengths at the high signal-to-noise ratio (SNR) region. In addition to the standard short cycle removal methods, we consider

two new optimization criteria unique to non-binary codes. for each individual short cycle, we assign non-binary values to the associated non-zero elements in the parity-check matrix in order to optimise the decoding performance. for overlapped short cycles, the non-zero elements are assigned by first improving the minimum distance and then optimising the decoding performance. The additional optimization requires only moderate increase in the code design complexity. Simulations results are used to demonstrate the effectiveness of the proposed method.

D.5.5 Outer code design for serially concatenated continuous phase modulation with symbol-wise interleading

Researchers: M. fu and J. Ning

Serially concatenated continuous phase modulation (SCCPM) schemes containing a binary outer code are good candidates in achieving both high spectral efficiency and high power efficiency. Recently, it is shown that SCCPM systems with symbol-wise interleaving can achieve substantial improvement in BER performance over their counterparts with bit-wise interleaving. The modulation index and the duration of the shape function in the modulation are important parameters for spectral efficiency. Conventionally, the outer code with order-2 is assumed to be superior to those with higher orders despite of the modulation index and the duration. This work considers symbol-wise interleaved SCCPM schemes and optimises the outer codes using extrinsic information transfer (EXIT) chart for different modulation indices and the durations. Also an additional intra-symbol interleaving is used to achieve the best decoding performance. Simulation results are used to show the superior performance of the proposed schemes.

D

Page 59: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control58

Program Goals:

The Bayesian Learning Program comprises researchers from Statistics, Engineering and Applied Mathematics backgrounds, reflecting the strong interdisciplinary nature of the Centre. In 2008 the Program focused on the following four themes in fundamental and applied research:

n New ways of describing data Topics: Nonparametrics, Priors, Use of expert data, Analysis of images

n New ways of modelling data Topics: Mixtures, Meta-analysis, Bayesian Networks

n New computational methods Topics: PMC, ABC, IS, Hybrid methods, Genetic algorithms

n New applications Topics: Biosecurity, Environment, Bioinformatics, Mining, Mapping, Health

The Bayesian Learning Program is located primarily at QUT Brisbane, led by CI Mengersen. It has inspired the formation of the active BRAG (Bayesian Research and Applications Group) at QUT, which comprises approximately a dozen PhD Students and the same number of postdocs and research associates.

Major research activities in 2008 included development of theoretical, computational and applied Bayesian methods and models, publication of outputs in the form of journal articles and conference papers, conduct of an international workshop and professional short courses, participation in national and international conferences, hosting of international visitors, visits to international laboratories, supervision of postgraduate Students and collaboration with other members of CDSC. These activities are detailed below.

Program Participants

Program Leader: Professor Kerrie Mengersen

Academic Researchers: Dr Darfiana Nur, Professor Tony Pettitt, Dr Rob Reeves, Professor Ian Turner

Research Associates: Dr Ross McVinish, Dr Mark Griffin, Dr Jaime Peters, Dr Paula Lennon; Dr David Allingham, Dr frank Tuyl

Continuing PhD Students: Kate Lee, Christoper Oldmeadow, Sandra Johnson (articulated from Research Masters)

Completed PhD Students: Dr Trevor Moffiet (graduated), Darren Wraith (submitted)

E. STATISTICS

Kerrie MengersenQUT Node Program Leader

EBAYESIAN LEARNING –

QUT NODE

Page 60: 2008 - University of Newcastle

2008 ANNUAL REPORT 59

E.1 PARAMETRIC BAYESIAN MODELLING

Project Leader: K. Mengersen

Researchers: C. Alston, M. Griffin, R. McVinish, J. Peters, J. Rayner, D. Wraith (Student), f. Tuyl (Student), J. Kelly (Honours Student), S. Moynihan (Research Assistant)

External Academic Collaborators: J. Tso (Visiting Student, Auckland University)

E.1.1 Spatio-temporal mixture models

Researchers: C. Alston, K. Mengersen and D. Wraith

Dr Alston, Mr Wraith and Professor Mengersen continue to study Bayesian spatio-temporal mixture models. The research focused on the characterization of priors for mixture components based on previous time-points and the hierarchical placement of these priors. The latter induces a different degree of smoothing and flexibility of description of patterns in the data. This work was written up by Darren Wraith and submitted as a PhD thesis in 2008. Two journal articles have also been submitted. As a result of this work, Darren was offered and has taken up a postdoctoral research position at the Universite Dauphine, Paris, france.

E.1.2 Image classification

Researchers: C. Alston, K. Mengersen and S. Moynihan

In collaboration with medical researchers, Dr Alston, Professor Mengersen and Mr Sean Moynihan continued research into Bayesian spatio-temporal mixture models for the problem of estimating volume of cement in hip replacements, based on (2-dimensional) CAT scan images and allowing for artifact.

Research also continued into the application of Baysian mixture models for classification of tissue components in a CAT scan, and comparison of these components across the body and before-after treatment.

This work motivated an investigation into the effect of using binned data on these models and corresponding analyses. A solution comprising a resampling step and an accept-reject algorithm has been developed and successfully demonstrated to provide superior classification outcomes.

E.1.3 Meta-analysis

Researchers: K. Mengersen and J. Peters

Dr Jaime Peters and CI Mengersen continued research in Bayesian models for multivariate and repeated measures meta-analysis. This work has led to the acceptance of two journal papers and the submission of a further two papers. These papers address the issues of alternative Bayesian models for repeated measures data, Bayesian updating in meta-analysis models, combining results adjusted for different confounders, and reasons for selective reporting of adjusted estimates in observational epidemiological studies.

E.2 BAYESIAN NONPARAMETRICS

Project Leader: J.H. Braslavsky

Researchers: J.H. Braslavsky and R. McVinish

This project, led by Dr Ross McVinish and Dr Julio Braslavsky, was completed in 2008. The project focused on the estimation of minimum phase transfer functions. The aim of the project was to quantify model uncertainty based on weak prior knowledge, and to apply this to the estimation of the transfer function for robust controller design. The approach taken was to specify a spline prior for the magnitude function and use this to determine the transfer function. Estimation was via MCMC. This work has been written up as a journal article and has been submitted to Technometrics.

E.3 COMPLEX MODELS

Project Leader: K. Mengersen

Researchers: C. Alston, S. Johnson (Student), K. Mengersen, D. Nur and D. Wraith (Student)

External Academic Collaborators: J. Rousseau (france)

E.3.1 Bayesian Networks

Researchers: S. Johnson and K. Mengersen

PhD Student Sandra Johnson and CI Mengersen continued to pursue research into the methodology and application of Bayesian Networks (BN). Continuing applied projects included the identification of scientific and management factors affecting the initiation and growth of lyngbya in Moreton Bay (with Healthy Waterways in Qld), and the survival and successful relocation of wild cheetahs in Southern Africa (with cheetah conservation agencies in South Africa, Namibia and Botswana). New applied projects included the use of BN to identify risk pathways for plant biosecurity (with Chevron Pty Ltd), and to map and quantify complex systems in airports (with a range of Australian and international airport groups). The methodological issues addressed in this research included the development and use of time slices to represent dynamic and interacting features in a BN, and the integration of multiple BN. An example of the latter is the design and implementation of a method for successfully integrating a BN representing scientific factors influencing an (environmental) outcome with a BN representing management factors influencing this outcome, via a third party ‘risk hazard map’.

E

Page 61: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control60

E.3.2 Genetics

Researchers: K. Mengersen, D. Nur and J. Rousseau (france)

Dr Darfiana Nur (UN), Professor Judith Rousseau (france) and CI Mengersen continued their investigation into the sensitivity of prior specification of a Bayesian hidden Markov model for DNA sequence segmentation. The sensitivity analysis included a traditional approach, varying the prior distributions for base transition probabilities for each segment type and for the transition matrix of segment types. A sequence of Dirichlet priors was considered for the former and Dirichlet and mixture Dirichlet priors for the latter. A second approach was also examined and promoted, which employed importance sampling of an MCMC chain obtained from the traditional approach. The study revealed the feasibility and computational efficiency of this approach for comparing a large number of priors simultaneously in a more comprehensive sensitivity analysis.

E.3.3 Environment and Health

Researchers: C. Alston, K. Mengersen and D. Wraith

Dr Clair Alston, PhD Student Darren Wraith and CI Mengersen continued to apply the research on spatio-temporal mixture models, described above, to problems in environment and health. One such problem involves the characterization of the size and concentration of air pollution particles and their evolution through time. This required the formulation and evaluation of dynamic mixture models using a range of priors based on the previous temporal data. A second problem involves ‘bioregionalisation’ of an area, that is, the (fuzzy) separation of an area into homogeneous regions based on a range of biological and environmental variables.

A new project has commenced in collaboration with the NSW Department of Agriculture, to develop a Baysian three-dimensional model comprising space, time and depth, for modelling soil moisture as part of a global warming study.

E.4 BAYESIAN PRIORS

Project Leader: K. Mengersen

Researchers: K. Mengersen, S. Low Choy and A. James

CI Mengersen has undertaken joint research with colleagues Dr Sama Low Choy and Mr Allan James into methods for eliciting expert information and representing this information as priors in a Bayesian analysis. An indirect approach to elicitation was adopted, in which the expert provides opinion about a response based on a set of covariates, with a corresponding level of uncertainty about this opinion. The research has also focused on methods for combining priors elicited from multiple experts, and encoding expert opinion on skewed non-negative distributions. for the latter, the popular two-parameter gamma and lognormal distributions, as well as the three-parameter location-shifted lognormal and quantile-specified Davies distribution, were considered. Equations were derived for moment-matching approaches, each depending on a different set of summary statistics elicited from experts. The study highlighted the varying accuracy that can be achieved, depending on the encoding method (which summary statistics are used), the distributional choice and the expert.

A prototype software package, Elicitator, has been developed to assist in the elicitation of priors that can be defined in a geographic context. This work has been applied to the problem of identifying habitat suitability for rare and threatened species. The following figure illustrates the ‘Response Viewer’ in Elicitator, which shows site predictions and associated encoded responses for each covariate at one site in the habitat suitability study.

figure 37

E

Page 62: 2008 - University of Newcastle

2008 ANNUAL REPORT 61

E.5 COMPUTATIONAL METHODS

Project Leader: K. Mengersen

Researchers: K. Lee (Student), R. McVinish, K. Mengersen

External Academic Collaborators: C. Robert (Université Paris Dauphine, france)

Dr Ross McVinish, PhD Student Kate Lee, french collaborator Professor Christian Robert and CI Mengersen continued research into computational algorithms for Bayesian analysis.

E.5.1 Population Monte Carlo

A theoretical investigation was completed of the performance of one such algorithm, Population Monte Carlo (PMC), that is gaining in popularity as a static version of its more well-known sequential counterparts such as the Kalman filter. The basis of these algorithms is Importance Sampling (IS), and the essence of the PMC algorithm is to learn from experience, that is to adapt the IS based on previous IS proposals, in order to approximate the target distribution more rapidly and accurately. However, due to the novelty of the PMC, its performance in high dimensions typically encountered in Bayesian analysis is not well known.

This research study focused on examining performance in terms of precision of estimates, and derived results relating to the asymptotic variance of the estimate as a function of the dimension. The results demonstrated exponential growth in the asymptotic variance of the estimate with increasing dimension, and proposed a method for determining the scale of the optimal proposal, which is critical to the performance of the algorithm.

E.5.2 Hybrid MCMC methods

The development of efficient computational techniques for Bayesian analysis often involves the combination of existing techniques to compose so-called ‘hybrid’ algorithms. This study examined the performance of selected hybrid algorithms: the delayed rejection algorithm, the pinball sampler, the Metropolis adjusted Langevin algorithm and the population Monte Carlo algorithm. Performance was evaluated in terms of statistical efficiency (including rate of convergence and mixing speed), applicability (including ease of set-up and flexibility of choice of component parameters and distributions) and implementation (such as the level of difficulty in coding, memory storage and computational demand). The study found that even if each component of a hybrid algorithm has individual strengths, they may not contribute equally or even positively when they are combined.

E.5.3 Computational of mixture distribution

Mixture models are being popularly used as statistical descriptions of a wide variety of phenomena. Although mixtures of t-distributions have proposed as appealing alternatives to the more common mixture of normal distributions, especially for modelling data with long tails or atypical observations, the computational properties of such mixtures are not yet well understood. This study provided a guide for applying the t-mixture model, addressed the sensitivity of the model to prior formulation, and examined the implementation of the t-mixture in the analysis of two important environmental problems: the characterization of tissue in a CAT scan image and the description of the concentration and size of particulate matter at a specified time period.

E.5.4 Importance sampling

In Bayesian analysis, it is often desirable to simulate multiple MCMC chains in order to assess convergence to the target distribution as well as accuracy and precision of the resultant estimates. A severe constraint in this process is the length of time that it may take to simulate multiple chains, particularly for complex high-dimensional problems. In this case a possible alternative is to simulate a single chain and then take importance samples from this chain. Theoretical and empirical investigations focused on the approximating properties of the estimated statistic of interest H'(x) to H(x) based on the IS chain and the MCMC chain, respectively. In particular, if ’≤M for some constant M then H’(x) has the same ergodic properties as the usual MCMC estimator.

E

Page 63: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control62

Program Goals:

The group will develop statistical theory and its application to the efficient and effective interpretation of data for the advancement of knowledge in all branches of learning and enterprise. This will be achieved through research into basic statistical theory and through research to solve particular problems, often in collaboration with researchers from other areas.

The group’s mission is to provide appropriate and innovative solutions and contributions to the development of statistical theory, modelling, analysis and reporting with a primary focus on application. The group will become known for its breadth and depth of applications and knowledge, accessibility, professionalism and friendliness, and successful collaboration with researchers and the professional community across many fields.

The group will fulfil the above through the members’ established research interests and outcomes, continued individual innovation and self-development, collaborative efforts with both intra – and inter-institutional researchers and the community, involvement with community and professional societies, continued journal publications, grant applications and peer-reviewing of journals’ submitted articles.

Program Participants:

Program Leader: John Rayner

Deputy Leader: Robert King

Research Associates: David Allingham, John Best, Peter Howley, Darfiana Nur, Elizabeth Stojanovski, Barrie Stokes

Continuing PhD Students: Phillip Lane, Paul Rippon, fatimah Almah Saaid

Honours Students: Megan ford, Ian Robinson

Project Visitors: Olivier Thas, Ghent University, Belgium

E.6 BAYESIAN ESTIMATION Of QUANTILE DISTRIBUTIONS

Project Leader: D. Allingham

Researchers: R.A.R. King and K.L. Mengersen

Use of Bayesian modelling and analysis has become commonplace in disciplines such finance, genetics and image analysis. Many complex data sets are collected which do not readily admit standard distributions, and often comprise skewed and kurtotic data. Such data is well-modelled by the very flexibly-shaped distributions of the quantile distribution family, whose members are defined by the inverse of their cumulative distribution functions and rarely have analytical likelihood functions defined. Without explicit likelihood functions, Bayesian methodologies such as Gibbs sampling cannot be applied to parameter estimation for this valuable class of distributions without resorting to numerical inversion. Approximate Bayesian computation provides an alternative approach requiring only a sampling scheme for the distribution of interest. This enables easier use of quantile distributions under the Bayesian framework. Parameter estimates for simulated and experimental data are presented.

STATISTICAL INfERENCE AND MODELLING –

UON NODE

E

John Rayner UoN Program Leader

Robert King UoN Deputy Program Leader

Page 64: 2008 - University of Newcastle

2008 ANNUAL REPORT 63

E.7 DNA SEGMENTATION ANALYSIS

Project Leader: D. Allingham

Researchers: R.A.R. King and K.L. Mengersen

The problem of DNA segmentation is to identify the location and structure of homogeneous segments of DNA, where the term ‘structure’ refers to the set of intra-segment nucleotide transition probabilities. Without knowing the segment boundaries, we wished to both differentiate one segment from another as well as estimate the true, underlying nucleotide transition probabilities of each segment. The model used was a hidden Markov model with change-points.

Approximate Bayesian Computation (ABC) was used to estimate distributions of model parameters, even though the model likelihoods were tractable (and therefore a candidate for such techniques as Markov chain Monte Carlo). Our aim, however, was to develop a likelihood-free approach, one which would be applicable to more general, more complex, intractable, models.

To complete earlier work, we developed a parallel version of ABC to take advantage of the high-performance computing facilities available to us. When applying ABC, it is a requirement to choose a summary statistic in order to compare output from the model to the observed data. When available, a sufficient statistic is chosen, but no such statistic was available for this model. In such a case, a statistic should be chosen which characterizes the features of the model being estimated.

To this end, we developed a statistic based on moving-average smoothed oligonucleotide occurrence vectors, which essentially provides point estimates of the nucleotide transition probabilities regardless of the boundary locations. This was then used to successfully fit both the boundary locations and the intra-segment nucleotide transition probabilities simultaneously.

Building on this success, we then began an investigation of parameter estimation for stochastic dynamical models with applications in ecology and disease modelling, using closely-related summary statistics.

Presentations of this material were given at the ASEARC conference, IASC2008 in Yokohama and to the Institute of Industrial Science, The University of Tokyo.

E

Page 65: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control64

0 200 400 600 800 1000 1200 14000

0.01

0.02

Location

Prob

abilit

y

200 250 300 350 4000

0.005

0.01

0.015

0.02

Boundary 1 location

Prob

abilit

y

800 850 900 950 10000

0.005

0.01

0.015

0.02

Boundary 2 location

Prob

abilit

y

4 5 6 7 80

0.01

0.02

0.03

0.04

0 2 4 6 80

0.01

0.02

0.03

0.04

0.05

−5 0 5 100

0.02

0.04

0.06

0.08

−1.5 −1 −0.5 0 0.5 10

0.01

0.02

0.03

0.04

0.05

figure 38: Distributions of segment boundary locations for simulated DNA sequence, obtained via ABC with smoothed oligonucleotide occurrence summary statistic.

figure 39: Decreasing the distance threshold in ABC improves fits, shown here for 4 parameters of the g-and-k distribution. Note the unimodality of the final fits.

E

Page 66: 2008 - University of Newcastle

2008 ANNUAL REPORT 65

E.8 ANALYSING AND REPORTING CLINICAL INDICATORS USING BAYESIAN HIERARCHICAL MODELS

Project Leader: P. Howley

Researchers: Conjoint A/Prof R.W.Gibberd, Mr Stephen Hancock (School of Medicine and Public Health, faculty of Health, The University of Newcastle), Osita Ezeh (M.Sc. (Statistics) Student)

Dr Howley’s research has focused on Bayesian hierarchical modelling and its application to foster quality improvement activity in health care, through the creation of improved methods for analysis and reporting of clinical indicator data. This relates well with his research into ‘performance measures’, which extends beyond the health care field.

His research has explored the potential for a new control chart based on Bayesian models to improve the monitoring of a hospital’s clinical indicators. He is currently supervising a Masters Student who is near to submitting on this topic. Dr Howley presented his collaborative work on the analysis and reporting of clinical indicator data for the Australian Council for Healthcare Standards as an invited speaker at the QUT and St Andrew’s Medical Institute’s collaborative industry/academic workshop, “Driving excellence in clinical outcomes: methods for monitoring and influencing change”, in Queensland, May, 2008. Attendees included representatives from Queensland Department of Health, QUT, UQ, USQ, St Andrew’s Medical Institute and other major Queensland hospitals.

figure 40: Peter Howley addresses the QUT and St Andrew’s Medical Institute’s collaborative industry/academic workshop.

E.9 BAYESIAN HIDDEN MARKOV MODEL fOR DNA SEQUENCES SEGMENTATION MODELLING

Project Leader: D. Nur

Researchers: K. Mengersen and R. McVinish,

External Academic Researchers: J. Rousseau (University of Paris-Dauphine, france) M.G. Nair (University Western Australia) Yan-Xia Lin (University of Wollongong) N.D Yatawara (Curtin University of Technology)

Bayesian inference procedures and algorithms have revolutionized the field of computational biology since early 1990s due to the development of computer intensive simulation based methods such as Markov Chain Monte Carlo (MCMC). The use of these methods has led to increasingly complex models being fitted in many situations. for DNA sequence segmentation, a DNA sequence can be thought of as the observed process which evolves independently or dependently given the unobserved Markov chain which locates the position of the segment type. The parameters in this model are the base/nucleotide transition probabilities within segment types and the transition probabilities between segment types. This project focuses on the various aspects of a Bayesian hidden Markov model describing homogeneous segments of DNA sequences including prior sensitivity analysis, simulation and estimation.

E

Page 67: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control66

E.11 HIERARCHICAL MODELLING

Project Leader: E. Stojanovski

Researchers: R. King, P. Lane (Student) and K.L. Mengersen

Bayesian methods of structural equation modelling are common in the social and health sciences, particularly where longitudinal modelling is of primary interest. Applications based on women’s health data have been investigated. findings were presented at the International Biometric Conference (IBC) in Dublin, 2008.

A research grant was awarded in 2008 which resulted in a preliminary assessment of the variation in alcohol-related harm across communities in New South Wales. A Bayesian approach to hierarchical analysis was investigated and the spatial relationship between the variables of interest will be assessed. The tendency for data to be influenced by surrounding values will be incorporated into the analyses. Preliminary findings were presented in the form of a paper at the Conference of the International Association for Statistical Computing (IASC) 2008 in Yokohama, December 2008.

figure 41: Histogram of data hypothesised to be normal, a fitted normal density (full line) and two improved densities. The dashed density is less complex than the dotted density but does not agree as well with the data.

EE.10 SMOOTH TEST Of GOODNESS Of fIT

Project Leader: J. Rayner

Researchers: J. Best (Conjoint) and P. Rippon (Student),

External Academic Collaborators: O. Thas (Ghent University, Belgium)

The principal focus of this project for 2008 was the writing of a second edition of Smooth Tests of Goodness of fit. This will be published by Wiley early in 2009 with co-authors Dr. John Best, a conjoint at Newcastle, and Dr. Olivier Thas from Ghent University in Belgium; Olivier visited in february to progress this work. Seminars were given by Rayner to Goulburn 6 (attended by researchers from the Australian Bureau of Statistics, CSIRO, the Universities of Wollongong, Newcastle and Western Sydney and other academics) in March, to colleagues at the Texas A&M University in June, and to the International Conference on Interdisciplinary Mathematical and Statistical Techniques IMST 2008/fIM XVI in Memphis, Tennessee, USA, May and also in June. The provocative title was “...my brother ... is an hairy man, but I am a smooth man.”

Rayner, Thas and Best are working towards a special issue of the Journal of Statistical Theory and Practice on Modern Goodness-of-fit. The issue will appear in mid-2009.

The problem of goodness of fit is to assess if a data set is consistent with a specified probability distribution or model. In the smooth approach the probability density function of the specified distribution is nested in a rich family of alternatives. Orthonormal functions are used to construct this family. Typically the test statistic is the sum of squares of components that are asymptotically independent and asymptotically standard normal. The components are readily interpreted and yield powerful focused tests, while the sum of squares of components yields powerful omnibus tests.

Recent work has shown that these ‘typical’ properties hold only for models from exponential families of densities. Techniques have been developed that produce convenient tests for arbitrary distributions provided enough moments exist. This required the development of orthonormal polynomials for arbitrary distributions. Another innovation has been the combining of smooth methods with modern model selection techniques to produce alternative models when the hypothesised model fails. Several papers have been published that demonstrate the implementation of these methods for particular distributions, such as the logarithmic and Laplace distributions.

Page 68: 2008 - University of Newcastle

2008 ANNUAL REPORT 67

F. DISTRIBUTED SENSING AND CONTROL

Program Goals:

In this program we aim to develop new areas of research in the area of distributed sensing and control. The work builds on our work in the other programs, . Our goal is to establish and foster collaborative research arrangements with other Schools within the University of Newcastle as well as with internationally renowned external researchers and industry in both fundamental and applied research. We are currently working in three main areas, Biomedical Science, Power Transformer Modelling and Robotics.

F.1 NETWORKED CONTROL

F.1.1 Analysis and design of networked control systems

Project Leader: G.C. Goodwin

Researchers: G.C. Goodwin, D.E. Quevedo and E.I. Silva

This project is also partly funded under an ARC Discovery Project.

Networked Control has emerged in recent years as a new and exciting area in systems science. The topic has many potential applications in diverse areas ranging from control of microrobots to biological and economic systems. The supporting theory is very rich and combines aspects of control, signal processing, telecommunications and information theory. Our work has focused on performance of network control systems which experience data rate constraints, delayed data or data loss. We have studied both coder design for fixed controller and joint coder controller design.

We have also studied control of MIMO LTI plants and explored the potential benefits of replacing a traditional diagonal non-networked control architecture with a networked full MIMO one. Diagonal terms of the networked MIMO architecture employ transparent links, whereas the off-diagonal terms use communication channels which are subject to signal-to noise ratio constraints. Within this setup, we have shown how to design LTI coding systems which optimise performance. Our analysis reveals that the achievable performance in the networked situation may become arbitrarily poor, if the signal-to-noise ratio constraints in the communication links are sufficiently severe. In these cases, traditional decentralized controller structures are preferable. We have limited our analysis to the two-input two-output case and have illustrated our results for networked control systems with bit-rate limited communication channels – see Goodwin, Silva and Quevedo (2008) in International Plenary/Keynote Addresses, Journal Papers and Conference Papers, Quevedo, Ahlén and Goodwin (2008) in Conference Papers.w

James Welsh Program Leader

Rick Middleton Deputy Program Leader

f

Page 69: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control68

F.2 POWER TRANSfORMER MODELLING

Project Leader: J.S. Welsh

Researchers: S.D. Mitchell (Student) and J.S. Welsh

F.2.1 Distributed power transformer model for partial discharge location

Researchers: S.D. Mitchell (Student) and J.S. Welsh

This project utilises a narrowband high frequency distributed transformer model to estimate partial discharge location. Here a narrow band of frequencies within the frequency response of the system, that exhibit resonant pole behaviour, is specifically targeted. It is proposed that the observed response is due to the interaction of residual winding inductance with capacitance to ground. This physical phenomenon is inherently distributed; hence regions within the frequency response related to this interaction will be dependent upon the input location. With this premise an algorithm that estimates the location of the partial discharge by iteratively comparing the proposed model at various locations within the winding with the observed partial discharge frequency response has been implemented. The algorithm was tested on a single phase of a 66kV/25MVA interleaved transformer winding where the partial discharge was injected via an oil immersed point-plane 7.5kV source.

figure 42: PD frequency response when injected at the Bushing.

f

Page 70: 2008 - University of Newcastle

2008 ANNUAL REPORT 69

F.2.2 The influence of inductive disparity on the observed frequency response of a distribution transformer

Researchers: S.D. Mitchell (Student) and J.S. Welsh

This project uses generic three phase models to illustrate how the frequency response analysis (fRA) data of distribution transformers with different vector group numbers will be affected by inductive disparity. The inductive disparity of a three phase transformer is a well documented property due to variation in the observed core reluctance of transformer windings on different phases. At low frequency this will result in a subtle inductance inequality between the centre and outer limb windings of the transformer core. Understanding the influence of this disparity is important when analysing the frequency response of the transformer. frequency Response Analysis data taken from different terminals on the same transformer will display significant variation at low frequencies due to the effects of this inductive disparity. Effects include differences in the self resonant frequency and levels of attenuation. This project facilitates a more comprehensive understanding of the observed responses – see Mitchell, Welsh and Phung (2008) in Conference Publications.

figure 43: Generic low frequency model representing fRA high voltage to low voltage terminal test of a Dyn connected transformer.

f

Page 71: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control70

F.2.3 The influence of complex permeability on the broadband frequency response of a power transformer

Researchers: S.D. Mitchell (Student) and J.S. Welsh

This project considers the effect of the complex permeability of a power transformer core over the entire frequency response analysis (fRA) test spectrum. Current work in the fRA area generally neglects the core beyond a 200kHz. This project demonstrates that the effective complex permeability is still significant at 1MHz and remains above unity at frequencies exceeding 15MHz. The project also demonstrates that for broadband small signal testing, such as fRA, the low field conditions induced by the injected signal result in a relative permeability that approaches the initial permeability of the core. This ensures that the relative permeability remains approximately constant over a large range of frequencies and will have a degree of independence with respect to the injection voltage source – see Mitchell, Welsh and Phung (2008) in Conference Publications.

f

figure 44: Example of a transformer’s effective permeability; a) u’ (real component), b) u’’ (imaginary component).

Page 72: 2008 - University of Newcastle

2008 ANNUAL REPORT 71

F.3 ROBOTICS RESEARCH

Project Leader: R.H. Middleton

Researchers: S. Chalup, R. King, R.H. Middleton, J.S. Welsh, N. Henderson (Student), J. Kulk (Student), S. Nicklin (Student), A. Wong (Student),

The Standard Platform League of RoboCup 2008 was won by the NUManoids, a collaborative team from the University of Newcastle, and the National University of Ireland.

RoboCup 2008 saw the introduction of the Aldebaran NAO as the new standard platform for the SPL League, replacing the discontinued Sony AIBO that has been used in previous years. The field was also changed, coloured beacons were removed, and realistic goals with nets were used. These changes are in accordance with RoboCup’s progressive movement towards replicating a human soccer match.

The transition to a new humanoid robot from a quadruped presents many new and interesting challenges. In particular, a new locomotion engine designed for a biped is required. The upright stance and improved camera provide a new perspective from which to locate and identify objects. furthermore, the field alterations mean that localisation is performed using only the field lines and the goals.

There was limited preparation time for this year’s competition. Hence, drawing on previous experience a simple solution, which worked very well in practice, was adopted. The centre strategy was to locate and gain possession of the ball as quickly as possible. The ball was then immediately disposed of. This strategy was very effective against slower robots with time consuming behaviours.

F.3.1 Automated colour calibration system using multivariate gaussian mixtures to segment colour space

Researchers: S. Chalup, N. Henderson and R. King

In this project we have developed a system for automating the time consuming task of manual colour calibration for a mobile robot. By converting a series of YUV images to HSI format and analysing histogram data it can be seen that there are distinct regions of colour space for each object colour class and that one dimension; hue, can be used to uniquely identify each colour class. Using an expectation maximisation (EM) algorithm to estimate the parameters of a Gaussian mixture model, it is proposed that the HSI colour space can be segmented and automatically labelled for the purpose of automatic colour calibration. This method has been applied to a Aldebaran NAO robot vision system that uses a ‘soft’ colour classification method to classify non-unique colour space. By reducing the colour labelling dimension to one and implementing soft classification principles, a reliable automatic calibration system was achieved – see Henderson, King and Chalup (2008) in Conference Publications.

F.3.2 Sound-scapes for robot localisation through dimensionality reduction

Researchers: S. Chalup and A. Wong

Sound-scapes similar to landscapes, are geometric representations of an objects’ relative positions in the real world. In this project we demonstrate how to obtain and use a sound-scape to assist the Aldebaran NAO with localisation. We have applied dimensionality reduction techniques such as statistical learning methods which include neural networks, support vector machines, the recent graph based approximation technique isometric feature mapping to extract the NAO’s field co-ordinate from its recorded acoustic data. Results obtained include visualisations of sound-scapes (robot’s positions on field) and positional accuracies of up to 80% – see Wong and Chalup (2008) in Conference Publications.

f

Page 73: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control72

F.4 SYSTEM IDENTIfICATION TOOLS fOR NEURON RESPONSE EVENTS

Project Leader: J.S. Welsh

Researchers: A.J. Rojas and J.S. Welsh

Biomedical Science is a research area of growing interest for the system theory community. It presents difficult and challenging problems. The data obtained from experiments requires processing and interpretation to obtain suitable models that can help explain the observed behaviour.

In this project we propose to study the modelling of electrical burst activity registered in neurons located in the spinal cord linked to the processing of pain signals. The recordings of the electrical burst activity are obtained using a method known as voltage clamping. This data is supplied by the group of researchers headed by Professor Alan Brichta, from the School of Biomedical Sciences.

It is known that the electrical bursts are caused by the diffusion of GABA and Glycine neurotransmitters. In particular we aim to investigate the use of time and frequency-domain identification tools to help differentiate between GABA and Glycine events. Currently this is performed using a pharmacological approach. By removing the need for pharmacological agents will allow the extension of recordings from an in vitro to an in vivo setting. Initial results are contained in the Technical Report DSC/BME/08/01.

F.3.3 A low power walk for the NAO robot

Researchers: J. Kulk and J.S. Welsh

Generally online walk pattern generators for humanoids are simplified, and don’t produce ideal gaits. Allowing the robot to ‘settle’ into a more natural gait through the modification of the low-level positional controller would provide significant benefits. In this project we attempt to achieve this, by limiting the power available to each motor in a humanoid, hence restricting how rigidly the joint can follow the generated walk pattern. This approach was evaluated by implementing the control modification on a humanoid robotics platform. The results show a significant improvement in walk speed, efficiency, and robustness. Moreover, the approach used here could be easily applied to any walk pattern generator, as the modification is in the low-level positional control – see Kulk and Welsh (2008) in Conference Publications.

f

Page 74: 2008 - University of Newcastle

2008 ANNUAL REPORT 73

G. MATHEMATICAL SYSTEMS THEORY

Program Goals:

The object of the program is to add to the expanding battery of mathematical knowledge that allows us to better understand both continuous and discrete dynamic systems exhibiting complex behaviour. In addition, expertise of members of the program in optimisation and linear and nonlinear analysis is exploited to aid the solution of problems driven by particular projects being undertaken by the Centre.

Throughout 2008 research proceeded on the following fronts.

G.1 BEYOND THE SPECTRUM

Project Leader: G. Willis

Researchers: G. Willis and J. Kimberley (Research Assistant, Newcastle).

External Academic Collaborators: f. Ghahramani (University of Manitoba, Winnipeg, Canada) C. Read (University of Leeds, United Kingdom) V. Runde (University of Alberta, Canada)

Various transforms that are important in applications of mathematics, such as the fourier and Laplace transforms and diagonalisation of matrices, convert a complicated multiplication in an algebra to a pointwise multiplication of functions. This technique relies on the existence of a space, sometimes called the maximal ideal space or the spectrum of the algebra, on which the functions are supported. However such a space does not always exist, for example, nilpotent matrices and Volterra integral operators cannot be represented as functions on a space, and in this situation the algebra is called radical.

This project aims to develop general techniques that may be used to analyse radical algebras, thus replacing spectral methods when they yield no information. Willis reported on this work at a Banach algebras conference held in honour of P.C. Curits Jr for the opening of the Philip C. Curtis Center for Mathematics and Teaching in March 2008.

Brailey Sims Program Leader

Jose De Doná Deputy Program Leader

George Willis Research Coordinator

Page 75: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control74

G.2 NONLINEAR ANALYSIS, OPTIMISATION AND fIXED-POINT THEORY

Project Leader: B.Sims

Researchers: J.M. Borwein, B.Sims, M.Bacak (Post Doctoral fellow), f.E. Castillo Santos (Student – supported by a CONACYT scholarship from the Mexican Government and employed as a part time Research Assistant), I. Searston (Student) and Matt Skerritt (Student).

External Academic Collaborators: S. Dhompongsa (Chiang Mai University, Thailand) K. Goebel (Marie-Curie Sklodowska University, Poland) H. Hudzyk (University of Poznan, Poland) M.A. Japon Pineda (University of Seville, Spain) W.A. (Art) Kirk (University of Iowa, USA) C. Lennard (University of Pittsburgh, USA) Gang Li (Yang Zhou University, China) E.L fuster (University of Valencia, Spain)

Equilibria for discrete and continuous dynamical systems correspond to fixed points of nonlinear maps on infinite dimensional function spaces. The solution of nonlinear optimisation and control problems lead to variational inequalities and thence to fixed points of related nonlinear operators. When the system is dissipative, these operators are typically nonexpansive with respect to an appropriate metric. Convergence and ergodic structure of orbits and various iterative schemes, such as those of Ishikawa, relate to the stability and long-term average behaviour of the system.

A principal goal is to further our understanding of nonexpansive and related types of mappings, with an emphasis on identifying readily applied, yet widely applicable, criteria that ensure the existence of fixed points for such maps together with effective algorithms by which they can be approximated. Special emphasis is given to the more difficult cases, where the underlying space lacks nice geometric structure such as that exhibited by, for example, a Hilbert space, or where convexity, or linear structure, is abscent.

figure 45

Page 76: 2008 - University of Newcastle

2008 ANNUAL REPORT 75

G.2.1 Alternating-projection and reflection algorithms

Researchers: J. Borwein, B. Sims and M. Skerritt

The performance of many alternating projection type algorithms is covered by a well established theory in the case of convex constraint sets. However, these same algorithms have been, and continue to be, used to successfully to ‘solve’ many core non-convex signal reconstruction problems in astronomy, physics, bioscience, genomics, geoscience, and medicine, but without real justification. Experience and computer experimentation reveals much about the likely behaviour in such situations, but underpinning theory is absent. Borwein, Sims and Skerritt are working to fill this void. Success will provide better methods for such reconstructions and, equally importantly, an understanding of why and when they work. This research on optimisation algorithms promises to improve the performance and quality of many practical signal reconstruction methods used by varied (Australian) industries from telecommunication to mining and by researchers in the digital arts and fields such as astronomy, physics, chemistry, bioscience, geoscience, engineering and health.

G.2.2 Analysis in the absence of linearity

Researchers: A. Kirk (USA), I. Searston and B. Sims

In situations where there is no intrinsic linear structure present (for example; certain instances of image reconstruction and aspects of pattern recognition and cognitive modeling) one must rely on non-algebraic (geometric) properties of the space to carry the analysis.

Recently it has been observed that certain geodesic (or Menger convex) metric spaces, in particular the so called CAT(0) spaces, provide a very general setting in which a rich analysis seems possible. W.A. (Art) Kirk has made important progress toward developing a fixed point theory for nonexpansive type mappings and multifunctions in this setting. We (Kirk, Searston and Sims) are working to extend this and to establish alternating projection type algorithms for solving ‘metrically convex’ feasibility problems.

G.2.3 Fixed points in the absence of weak compactness

Researchers: M.A. Japon Pineda (Spain), C. Lennard (USA), f.E. Castillo Santos and B. Sims

In mid 2006 Pie-Kee Lin obtained a seminal result which answered (unexpectedly) in the negative a widely investigated conjecture first raised by Sims nearly three decades ago. Lin showed that under an equivalent dual renorming, initially considered by Lennard and Dowling, the nonreflexive space X = 1

1 enjoyed the fixed point property for nonexpansive maps (fpp); that is, for every nonempty closed bounded convex subset C of X all nonexpansive mappings T:C➝C have a fixed point. Previously it had been conjectured this would only happen if X were reflexive. Maria A. Japon Pineda, Chris Lennard, Castillo Santos and Sims are successfully investigating what special features of this particular norm ensure it has the fpp, in order to establish the fpp for a wider class of renormings of 11 as well as carry the result to other spaces.

G.2.4 Semigroups of mappings

Researcher: Gang Li and B. Sims

The existence of (common) fixed points, together with associated asymptotic behaviour and ergodic theory for (semigroups of) nonlinear mappings have proved important in many areas, including the study of evolving systems, control theory and optimisation. Led by Gang Li, we have obtained very general demiclosedness principles, on not necessarily convex sets, and results of the three types for (commutative semigroups of) both certain Lipschitzian; nonexpansive, and non-Lipschitzian; asymptotically nonexpansive, mappings on a special class of metric spaces (including many Banach spaces). One paper has appeared and five more are in the process of being submitted.

G.2.5 Ultraproduct methods

Researchers: H. Hudzyk, W.A. Kirk and B. Sims,

Banach space ultra-products, and more recently ultra-products of metric spaces, represent a common meeting ground between standard and non-standard analysis, and have become powerful tools in both linear and nonlinear analysis. They have proved central to the fixed point theory of nonexpansive type mappings in the absence of strong geometric properties, such as normal structure, and are particularly suited to the analysis of attainment problems. By lifting the problem to an ultra-power approximate solutions become exact solutions (for example; an approximate eigenvalue of an operator corresponds to an eigenvalue of the lifted operator).

Refinements of these techniques and their application to a wide variety of problems is one of our major focuses.

We look forward to 2009 when the group will be joined by three new ongoing appointments: Jeffery Hogan (wavelet and harmonic analysis, image reconstruction methods); Xin-She Yang (modeling Non-porous Media with coupled systems of partial differential equations applied to geotectonics and pattern formation); Wadim Zudlin (special functions, analytic number theory).

figure 46: Reflect-reflect working in the case of an affine manifold and sphere

Page 77: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control76

PUBLICATIONS * Denotes international author. # Denotes publication not funded by CDSC.

BOOKS J.M. Borwein and J. Vanderwerff* Convex functions: Constructions, Characterizations and Counterexamples, Encyclopedia of Mathematics and Applications, Cambridge University Press.

#J. Borwein and K.Devlin* The Computer as Crucible: an Introduction to Experimental Mathematics, AK Peters, October 2008. ISBN-13: 978-1568813431.

#J.M. Borwein, E.M. Rocha* and J.f. Rodrigues* Communicating Mathematics in the Digital Era, AK Peters, September 2008. ISBN-13: 978-1568814100.

#J.M. Borwein and D.H. Bailey* Mathematics by Experiment: Plausible Reasoning in the 21st Century A.K. Peters Ltd, 2004, ISBN: 1-56881-136-5. Combined Interactive CD version 2006. Second expanded edition, September 2008.

S. Dhompongsa*, K. Goebel* and W.A. Kirk*, S. Plubtieng*, B. Sims and S. Suantai* (Editors), fixed point theory and its applications, Proceedings of the 8th International Conference held in Chiang Mai (Thailand), July 16--22, 2007, Yokohama Publishers, Yokohama, 2008. iv+230 pp. ISBN: 978-4-946552-31-1

G.C. Goodwin, f. Sobora and A. Bastiani A Virtual Laboratory for Control System Design. See www.virtual-laboratories.com

BOOKS IN PREPARATION/ TO APPEAR

J.C.W. Rayner, O. Thas* and D.J. Best Smooth Tests of Goodness of fit: with R Package (2nd ed.). Wiley. To appear.

CHAPTERS IN BOOKS

G.J. Adams “Making Advances in Mineral Exploration”, in Outcomes: Results of Research in the Real World, edited by Australian Research Council, pp.67-68, June 2008.

J.A. De Dona, f. Suryawan, M.M. Seron, and J. Levine* “A flatness-based iterative method for reference trajectory generation in constrained NMPC, Assessment and future Directions of Nonlinear Model Predictive Control”, Lecture Notes in Control and Information Sciences, Springer-Verlag. In press.

G.C. Goodwin, “Advanced Control: Surmounting Complexity”, in Outcomes: Results of Research in the Real World, edited by Australian Research Council, pp.111-113, June 2008.

#G.C. Goodwin, J. Østergaard, D.E. Quevedo and A. feuer* “A Vector Quantisation approach to Scenario Generation for Stochastic NMPC” in future Directions in Nonlinear Model Predictive Control, edited by D. M. Raimondo, f. Allgower and Lalo Magni, Springer Verlag, 2009.

N. Henderson, R. King and R.H. Middleton “An application of Gaussian mixtures: Colour segmenting for the four legged league using HSI colour space in RoboCup 2007: Robot Soccer World Cup XI” (in Lecture notes in Computer Science), Berlin / Heidelberg: Springer. http://dx.doi.org/10.1007/978-3-540-68847-1-23

T. Perez “Optimising the Performance of Marine Vessels”, in Outcomes: Results of Research in the Real World, edited by Australian Research Council, pp.40-41, June 2008.

Y.I. Yuz* and G.C. Goodwin “Robust Identification of Continuous-Time Systems from Sampled Data” Chapter in Continuous Time System Identification, edited by L. Wang and H. Garnier, Springer, March 2008.

PATENTS

A. J. fleming “Positioning System and Method” Provision Patent Application filed. Newcastle Innovation Ltd.

PLENARY AND KEYNOTE ADDRESSES

J.M. Borwein “future Challenges for Variational Analysis” Conference on Variational Analysis and Nonlinear Optimization, Sun-Yat-Sen University, Taiwan, November 2008.

J.M. Borwein “Some of my favourite Convex functions” 7th NZ-AustMS Joint Meeting, ANZMC2008, Christchurch, NZ, December 2008.

J.M. Borwein “future Challenges for Variational Analysis” Special Session on Nonlinear Optimization and Applications, 7th NZ-AustMS Joint Meeting, ANZMC2008, Christchurch, NZ, December 2008.

M. fu “Education of Control and Automation: Challenges and future” Deans Conference on Education of Control and Automation, China, April 2008.

#G.C. Goodwin, E.I. Silva and D.E. Quevedo “Analysis and Design of Networked Control Systems” Chinese Control and Decision Conference, Yantai, China, 2nd July 2008.

G.C. Goodwin, J. Ostergaard, D.E. Quevedo and A. feuer* “A Vector Quantisation Approach to Scenario Generation for Stochastic NMPC” International Workshop on Assessment and future Directions of Nonlinear Model Predictive Control, Pavia, Italy, 5-9 September 2008.

K. Mengersen, Keynote speaker, International Conference on Monte Carlo Methods (MCMSki), Italy, 9-11 July 2008.

J.C.W. Rayner “... my brother ... is an hairy man, but I am a smooth man”, International Conference on Interdisciplinary Mathematical and Statistical Techniques, IMST 2008/fIM XVI in Memphis, Tennessee, USA, May 2008.

Page 78: 2008 - University of Newcastle

2008 ANNUAL REPORT 77

INVITED PRESENTATIONS

#J.M. Borwein “Digitally-assisted Discovery and Proof” Special Session on University Mathematics Education, Teaching and Learning, 7th NZ-AustMS Joint Meeting, ANZMC2008, Christchurch, December 2008.

J.C.W. Rayner, “To Goulburn 6” Presented to researchers from the Australian Bureau of Statistics, CSIRO, the Universities of Wollongong, Newcastle and other academics) March 2008

J.C.W. Rayner, “... my brother ... is an hairy man, but I am a smooth man.” Presented at Texas A&M University in June 2008.

K. Mengersen Invited Public Seminar, University of Southern Queensland, 31 March 2008

T. Moffiet “The estimation of forest foliage cover from LANDSAT spectral imagery June 2008” Presented to Queensland Remote Sensing Centre at the Queensland Department of Natural Resources and Water

B. Sims “Signal Reconstruction, fienup’s algorithm and the Douglas-Rachford (Lions-Mercier) algorithm” Special Workshop of the Hashemite University, Jordan, December 2008.

G. Willis “Radical Banach Algebras” Philip C. Curtis Center for Mathematics and Teaching Conference, University of California, Los Angeles, March 2008.

#G. Willis “Totally disconnected groups” Oberwolfach Workshop Buildings: Interactions with Algebra and Geometry, January 20 – 26, 2008, Germany.

G.C. Goodwin “ProcessACT: A Tool for Advanced Control System Design and Implementation” Graeme Clark Research Outcomes forum, Parliament House, Canberra, 18 June 2008.

S.O.R. Moheimani “Control and Dynamics of Micro and Nanosystems” Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

JOURNAL PAPERS

J.C. Agüero and G.C. Goodwin “Identifiability of errors in variables dynamic systems”, Automatica, Vol.44, No.2, pp.371-382, february 2008.

D. Allingham, R.A.R. King and K.L. Mengersen “Bayesian estimation of Quantile distributions”, Statistics and Computing online 25 July 2008, 13 pages, doi:10.1007/s11222-008-9083-x.

S. Aphale, B. Bhikkaji and S.O.R. Moheimani “Minimizing scanning errors in piezoelectric stack-actuated nanopositioning platforms”, IEEE Transactions on Nanotechnology, Vol.7, No.9, pp.79-90, January 2008

S. Aphale, S. Devasia* and S.O.R. Moheimani High-bandwidth control of a Piezoelectric nanopositioning stage in the presence of plant uncertainties Nanotechnology, Vol.19, No.12, 9 pages, March 2008

D.H. Bailey*, J.M. Borwein, D.M. Broadhurst* and L. Glasser* “Elliptic integral representation of Bessel moments”, J. Phys. A: Math. Theor, Vol.41, pp.5203-5231. DOI 205203 (IoP Select). (http://arxiv.org/abs/0801.0891)

#R. Baillie*, D. Borwein* and J. Borwein “Some sinc sums and integrals”, American Math. Monthly, Vol.115, No.10, pp.888.901, October 2008.

D.J. Best, J.C.W. Rayner and O. Thas* “X2 and its components as tests of normality for grouped data”, Journal of Applied Statistics, 35(5), 481-492.

D.J. Best, J.C.W. Rayner and O. Thas* “Tests of fit for the logarithmic distribution”, Journal of Applied Mathematics and Decision Sciences. Article ID 463781, 8 pages, doi:10.1155/JAMDS/2008/463781.

D.J. Best, J.C.W. Rayner and O. Thas* “Comparison of some tests of fit for the Laplace distribution”, Computational Statistics and Data Analysis, No.52, Vol.12, pp.5338-5343, December 2008.

P. Butow, J. Cockburn, A. Girgis, D. Bowman, P. Schofield, C. D’Este, E. Stojanovski and M. Tattersall “Increasing oncologists’ skills in eliciting and responding to emotional cues: evaluation of a communication skills training program”, Psycho-Oncology, Vol.17, pp.209-218.

D. Borwein*, J.M. Borwein and R.E. Crandall* “Effective Laguerre asymptotics”, SIAM J. Numer. Anal., No.6, pp.3285-3312, 2008.

#D. Borwein*, J.M. Borwein and O-Yeat Chan* “The evaluation of Bessel functions via exp-arc integrals”, J. Math. Anal. Appl., No.341, pp.478-500, 2008.

#J. Borwein, A. Guirao*, P. Hajek* and J. Vanderwerff* “Uniformly convex functions on Banach spaces”, Proc AMS., Vol.137, pp.1081-1091, 2009.

#J.M. Borwein and O-Yeat Chan* “Uniform inequalities for the incomplete complementary Gamma-function”, Mathematical Inequalities and Applications, April 2008.

#J.M. Borwein and D.R. Luke* “Dynamics of a Ramanujan-type continued fraction with cyclic coefficients”, Ramanujan Journal, Vol.16, pp.285-304.

#J.M. Borwein and B. Salvy* “A proof of a recursion for Bessel moments”, Experimental Mathematics, Vol.17, pp. 223-230, 2008

#J.M. Borwein, I. J. Zucker* and J. Boersma* “The evaluation of character Euler double sums”, Ramanujan Journal, Vol.15, pp.377-405, 2008.

#J.M. Borwein “Hilbert inequalities and Witten Zeta-functions”, American Math Monthly, Vol.115, No.2, pp.125-137,

B. Bhikkaji and S.O.R. Moheimani “Integral resonant control of a piezoelectric tube actuator for fast nano-scale positioning”, IEEE/ASME Transactions on Mechatronics, Vol.13, No.5, pp.530-537, October 2008

Page 79: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control78

J. Chen*, S. Hara, L. Qiu, R.H. Middleton “Best achievable tracking performance in sampled-data control systems” IEEE Transaction on Automatic Control, Vol.53, No.11, pp2467-2479, November 2008.

G.M. Clayton*, S. Tien, A.J. fleming, S.O.R. Moheimani and S. Devasia* “Inverse-feedforward of charge-controlled piezopositioners”, Mechatronics, Vol.18, No.5-6, pp.273-281, June 2008

D.f. Coutinho*, M.fu, A. Trofino* and A. Danes* “L2-gain analysis and control of uncertain nonlinear systems with bounded disturbance inputs”, International Journal of Robust and Nonlinear Control, Vol.18, No.1, pp.88-110, January 2008.

#M.S. Derpich, E.I. Silva, D.E. Quevedo and G.C. Goodwin “On Optimal Perfect Reconstruction feedback Quantisers”, IEEE Transactions on Signal Processing, Vol.56, No.8 Part 2, pp.3871-3890, Aug, 2008.

C.E. de Souza*, K.A. Barbosa* and M. fu “Robust filtering for uncertain linear discrete-time descriptor systems”, Automatica, Vol.44, No.3, pp.792-798, March 2008.

#A.C. Eberhard and J.M. Borwein “Second order cones for maximal monotone operators via representative functions” Setvalued Analysis Special issue for Boris Mordukhovich, Vol.16, pp.157-184, 2008.

A.J. fleming and K.K. Leang* “Simulation of dynamics-coupling in piezoelectric tube scanners by reduced order finite element models”, “Ultramicroscopy”, Vol.108, No.12, pp.1551-1557, November, 2008.

A.J. fleming, A. Wills and S.O.R. Moheimani “Sensor fusion for improved control of piezoelectric tube scanners”, IEEE Transactions on Control Systems Technology, Vol.16, No.6, pp.1265-1276, November 2008.

G.C. Goodwin, J.-C. Aguero, J.S. Welsh, G.J. Adams, J.I. Yuz*, and C.R. Rojas “Robust identification of process models from plant data”, Journal of Process Control, Vol.18, No.9 pp.810-820, September 2008.

#G.C. Goodwin, D.E. Quevedo and E.I. Silva “Architectures and coder design for networked control systems”, Automatica, Vol.44, No.1, pp.248-257, January 2008.

G.C. Goodwin, M.M. Seron and D.Q. Mayne* “Optimization opportunities in mining, metal and mineral processing”, IfAC Annual Reviews in Control, Vol.32, No.1, pp.17-32, April 2008.

H. Haimovich* and M.M. Seron “Multivariable quadratically-stabilizing quantisers with finite density”, Automatica, Vol.44, Nr.7 , pp.1880-1885, July 2008.

M. Haynes, K.L. Mengersen and P. Rippon “Generalized control charts for non-normal data Using g-and-k distributions. Communications in Statistics: Simulation and Computation, Vol.37, No.9, pp.1881-1903.

P. Howley “Keeping it real, keeping them interested and keeping it in their minds”, Journal of Statistics Education, Vol.16, No.1, http://www.amstat.org/publications/jse/v16n1/howley.html

R.A.R. King and H.L. MacGillivray “fitting the generalized lambda distribution with location and scale-free shape functionals”, American Journal of Mathematical and Management Sciences, Vol.27, pp.441-460.

E.J. Kofman*, M.M. Seron and H. Haimovich* Control design with guaranteed ultimate bounds in perturbed systems, Automatica, Vol.44, Nr.7, pp.1815-1821, July 2008.

C. Løvaas, M.M. Seron and G.C. Goodwin “Robust output-feedback model predictive control for systems with unstructured uncertainty”, Automatica, Vol.44, No.8, pp1933-1943, August 2008.

S. Low Choy, K. Mengersen and J. Rousseau* “Encoding expert opinion on skewed non-negative distributions”, J. Applied Probability and Statistics, Vol.3, pp.1-21.

#J. Maess*, A.J. fleming and f. Allgöwer* “Simulation of dynamics-coupling in piezoelectric tube scanners by reduced order finite element models” Review of Scientific Instruments , Vol, 79, 015105, pp.1-9, January 2008.

I.A. Mahmood, S.O.R. Moheimani and B. Bhikkaji “Precise tip positioning of a flexible manipulator using resonant control”, IEEE/ASME Transactions on Mechatronics, Vol.13, No.2, pp.180-186, April 2008

R. McVinish, J. Rousseau* and K. Mengersen “Bayesian goodness-of-fit testing with mixtures of triangular distributions”, Scandinavian Journal of Statistics, Digital Object Identifier (DOI) 10.1111/j.1467-9469.2008.00620.x, October 2008.

D.E. Miller* and R.H. Middleton “The limitations of trajectory tracking”, IEEE Transactions on Automatic Control, Vol.53, No.11, pp.2586-2601, November 2008.

S.O.R. Moheimani and Y.K. Yong “Simultaneous sensing and actuation with a piezoelectric tube scanner”, Review of Scientific Instruments, Vol.79, No.7, pp.1-5, July 2008

S.O.R. Moheimani Invited Review Article: “Accurate and fast nanopositioning with piezoelectric tube scanners: Emerging trends and future challenges”, Review of Scientific Instruments, Vol.79, No.7, Article Number 071101, 11 pages, July 2008

D. Nur, D. Allingham, J. Rousseau*, K.L. Mengersen and R. McVinish “Bayesian analysis of DNA sequences segmentation: A prior sensitivity analysis”,Computational Statistics and Data Analysis, doi:10.1016/j.csda.2008.07.007.

D. Nur, J. Rousseau*, D. Allingham, R. McVinish and K. Mengersen “Bayesian hidden Markov model for DNA sequence segmentation: a prior sensitivity analysis”,Computational Statistics and Data Analysis, Online doi:10.1016/j.csda.2008.07.007.

D. Nur, M.G. Nair and N.D. Yatawara* “Adaptive estimation in Smooth Threshold AR(1) models”, Journal Statistical Theory and Practice, Vol.2, No.1, pp.83-94.

R. O’Leary, J. Murray, S. Low Choy and K. Mengersen “Expert elicitation for Bayesian classification trees”, J. Applied Probability and Statistics, Vol.3, pp.95-106.

Page 80: 2008 - University of Newcastle

2008 ANNUAL REPORT 79

R.A. O’Leary, S. Low Choy, J. Murray, M. Kynn, R. Denham, T. Martin and K. Mengersen “Comparison of three expert elicitation methods for logistic regression on predicting the presence of the threatened brush-tailed rock-wallaby Petrogale penicillata”, Environmetrics, Online DOI: 10.1002/env.935.

E. Pereira*, S.O.R. Moheimani and S. Aphale “Analog implementation of the integral resonant control scheme”, Smart Materials and Structures, Vol.17, No.6, December 2008

T. Perez and T.I. fossen* “A derivation of high-frequency asymptotic values of 3D added mass and damping based on properties of The Cummins’ Equation”, Journal of Maritime Research, Spain. Vol.5, No 1, pp 65-77.April,

T. Perez and T.I. fossen* “Time-domain vs. frequency-domain methods for identification of fluid memory parametric models of marine structures”, Modeling Identification and Control (MIC), Norwegian Research Bulletin, Trondheim, Norway. Vol.29 No.1 p.1-19.

T. Perez and T.I. fossen* “Joint identification of infinite-frequency added mass and fluid-memory models of marine structures”, Modeling Identification and Control (MIC), Norwegian Research Bulletin, Trondheim, Norway. Vol.29, No.3, pp.93-102

T. Perez and G.C. Goodwin “Constrained predictive control of ship fin stabilizers to prevent dynamic stall”, Control Engineering Practice, Vol.16, No.4, pp.482-494, April 2008.

#D.E. Quevedo, E.I. Silva and G.C. Goodwin “Control over unreliable networks affected by packet erasures and variable transmission delays”, IEEE Journal on Selected Areas in Communications: Special Issue on Control and Communications, Vol.26, No.4, pp.672-685, April 2008.

#D.E. Quevedo, E.I. Silva and G.C. Goodwin “Subband coding for networked control systems” International Journal of Robust Nonlinear Control, Vol.18, pp.1-20, November 2008.

J.C.W. Rayner, O. Thas* and B. DeBoeck* “A generalised Emerson recurrence relation”, Australian and NZ Journal of Statistics, Vol.50, No.3, pp.235-240.

A.J. Rojas, J.H. Braslavsky and R. H. Middleton “fundamental limitations in control over a communication channel”, Automatica, Vol.44, No.12, pp.3147-3151, December 2008.

A.J. Rojas, J.H. Braslavsky and R. H. Middleton “Channel signal-to-noise ratio constrained feedback control: Performance and robustness” IET Control Theory and Applications, Vol.2, No.7, pp.595-605, July 2008.

#C.R. Rojas, J.C. Agüero, J.S. Welsh and G.C. Goodwin “On the equivalence of least costly and traditional experiment design for control”, Automatica, Vol.44, No.11, pp.2706-2715, November 2008.

A. Sebastian, A. Pantazi, S.O.R. Moheimani, H. Pozidis and E. Eleftheriou* “Achieving sub-nanometer precision in a MEMS storage device during self-servo write process”, IEEE Transactions on Nanotechnology, Vol.7, No.5, pp.586-595, September 2008.

M.M. Seron, X.W. Zhuo, J.A. De Doná and J.J. Martinez* “Multisensor switching control strategy with fault tolerance guarantees”, Automatica, Vol.44, No.1, pp..88-97, January 2008.

S. Snodgrass, D. Rivett, V. Robertson and E. Stojanovski “forces applied to the cervical spine during posteroanterior mobilization’, Journal of Manipulative and Physiological Therapeutics, Vol.32, No.1, pp.72-83.

R. Taghipour*, T. Perez and T. Moan* “Hybrid frequency-time domain models for dynamic response analysis of marine structures”, Ocean Engineering, Vol.35, pp.685-705.

f. Tuyl, R. Gerlach and K.L. Mengersen “A comparison of Bayes-Laplace, Jeffreys and other priors: the case of zero events”, The American Statistician, Vol.62, No.1, pp.40-44.

f. Tuyl, R. Gerlach and K.L. Mengersen “Inference for proportions in a 2(2 contingency table: HPD or not HPD?”, Biometrics, Vol.64, No.4, pp.1293-1296.

R.A. Walsh, f. Tzelepis and E. Stojanovski “Australian pension funds and tobacco investments: promoting ill health and out-of-step with their members”, Health Promotion International, Vol.23, pp.35-41.

R.A. Walsh, C. Paul, f. Tzelepis, E. Stojanovski and A. Tang “Is government action out-of-step with public opinion on tobacco control? Results of a New South Wales population survey”, Australian and New Zealand Journal of Public Health, Vol.32, No.5, pp.3482:488.

A. Wills, D. Bates, A.J. fleming, B. Ninness and S.O.R. Moheimani “Model predictive control applied to constraint handling in active noise and vibration control” IEEE Transactions on Control Systems Technology, Vol.16, No.1, pp.3-12, January 2008

D. Wraith and K. Mengersen “A Bayesian approach to assess interaction between known risk factors: the risk of lung cancer from exposure to asbestos and smoking”, Statistical Methods in Medical Research, Vol.17, pp.171-189.

J. Zheng and M. fu “Nonlinear feedback control of a dual-stage actuator system for reduced settling time”, IEEE Transactions on Control Systems Technology, Vol.16, No.4, pp.717-725, July 2008.

J. Zheng, M. fu, Y. Wang*, and C. Du* “Nonlinear tracking control for a hard disk drive dual-stage actuator system”, IEEE/ASME Transactions on Mechatronics, Vol.13, No.5, pp.510-518, Oct. 2008.

J. Zheng, Y. Guo*, M. fu, Y. Wang*, and L. Xie* “Development of an extended reset controller and its experimental demonstration” , IET Control Theory and Applications, Vol.2, No.10, pp.866-874, October 2008.

Page 81: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control80

JOURNAL PAPERS ACCEPTED fOR PUBLICATION

M. Alamir*, J.S. Welsh and G.C. Goodwin “Redundancy vs. multiple starting points in nonlinear systems related inverse problems”. Accepted for publication in Automatica, Vol.45, No.4, April, 2009.

D. Bailey* and J. Borwein “Highly parallel, high precision integration”. Accepted for publication in International Journal of Computational Science and Engineering.

D.H. Bailey*, J.M, Borwein and R.E. Crandall* “Resolution of the Quinn-Rand-Strogatz constant of nonlinear physics”. Accepted for publication in Experimental Mathematics.

#U. Baumgartner, M. Laca*, J. Ramagge and G. Willis “Hecke algebras from groups acting n trees and HNN-extensions”. Accepted for publication in Journal of Algebra.

U. Baumgartner, G. Schlichting* and G. A.Willis “Geometric characterization of flat groups of automorphisms”. Accepted for publication in Groups, Geometry and Dynamics.

J. Borwein, A. Guirao*, P. Hajek* and J. Vanderwerff*, “Uniformly Convex functions on Banach Spaces” . Accepted for publication in Proc AMS., Vol.137, pp.1081-1091, 2009). [D-drive Preprint 340].

#J.M. Borwein and W. B. Moors* Stability of closedness of cones under linear mappings”. Accepted for publication in J. Convex Analysis.

#J.M. Borwein and J. Vanderwerff* “Differentiability of conjugate mappings”. Accepted for publication in J. Convex Analysis

#J. Borwein, N. Calkin* and D. Manna* “Euler and Boole summation revisited”. Accepted for publication in Amer. Math Monthly, April 2009.

#J.M. Borwein and C. Hamilton* “Symbolic convex analysis: Algorithms and examples”. Accepted for publication in Mathematical Programming, Vol.116, pp.17-35, 2008.

#M. fu and L. Xie* “finite-level quantisation feedback control for linear systems”. Accepted for publication in IEEE Transactions on Automatic Control.

#M. fu and L. Xie* “Quantised feedback control for linear uncertain systems”. Accepted for publication in International Journal of Robust and Nonlinear Control.

#H.Glöckner* and G.A. Willis, “Classification of the simple factors appearing in composition series of totally disconnected contraction groups”. Accepted for publication in J. Reine Angew. Math.\/.

Y. Guo*, Y. Wang*, L. Xie*, and J. Zheng “Stability analysis and design of reset systems: Theory and an application”. Accepted for publication in Automatica.

S. Low Choy, R.A. O’Leary and K. Mengersen “Elicitation by design in ecology: using expert opinion to inform priors for Bayesian statistical models”. Accepted for publication in Ecology.

K. Lee, K.L. Mengersen, J.-M Marin* and C.P. Robert* “Bayesian inference on mixtures of distributions”. Accepted for publication in Platinum Jubilee of the Indian Statistical Institute.

J.B. Mare, J.A. De Dona, M.M. Seron, H. Haimovich* and J. Ramagge. “When does QP yield the exact solution to constrained NMPC?”. Accepted for publication in International Journal of Control.

D. Marelli, K. Mahata and M. fu “Linear LMS compensation for timing mismatch in time-interleaved ADCs”. Accepted for publication in IEEE Transactions on Circuits and Systems.

D.Marelli and I. Raeburn “Proper actions which are not saturated”. Accepted for publication in Proceedings of the American Mathematical Society.

R. McVinish and K. Mengersen “ Semiparametric circular statistics”. Accepted for publication in Computational Statistics and Data Analysis.

D. Nur, D. Allingham, J. Rousseau*, K.L. Mengersen and R. McVinish “Bayesian hidden Markov model for DNA sequence segmentation: a prior sensitivity analysis”, Accepted for publication in Computational Statistics and Data Analysis.

J. Peters and K. Mengersen “Selective reporting of adjusted estimates in observational epidemiology studies: reasons and implications for meta-analyses”. Accepted for publication in Evaluation and the Health Professions.

J. Peters and K. Mengersen “Repeated measures in meta-analysis”. Accepted for publication in J.Evaluation in Clinical Practice.

J.C.W. Rayner, D.J. Best and O. Thas* “Generalised smooth tests of goodness of fit”. Accepted for publication in the Journal of Statistical Theory and Practice, Special Issue on Modern Goodness of fit Methods. eds. J.C.W. Rayner, O. Thas and D.J. Best, Greensboro, North Carolina: Grace Scientific Publishing.

J.C.W. Rayner and A.M. Carolan “Partially parametric testing”. Accepted for publication in the Journal of Statistical Theory and Practice, Special Issue on Modern Goodness of fit Methods. eds. J.C.W. Rayner, O. Thas* and D.J. Best, Greensboro, North Carolina: Grace Scientific Publishing.

C. Rojas, J.-C. Agüero and J.S. Welsh “fundamental limitations on the variance of parametric models”. Accepted for publication in IEEE Transactions on Automatic Control, Vol.54, No.2 february, 2009.

E. Ruth*, O.N. Smogeli*, T. Perez, and A.J. Sørensen* “Anti-spin thrust allocation for marine vessels”. Accepted for publication in IEEE Transactions on Control System Technology.

O. Thas*, J.C.W. Rayner, D.J. Best and B. DeBoeck* “Informative statistical analyses using smooth goodness of fit test”. Accepted for publication in the Journal of Statistical Theory and Practice, Special Issue on Modern Goodness of fit Methods. eds. J.C.W. Rayner, O. Thas and D.J. Best, Greensboro, North Carolina: Grace Scientific Publishing.

Page 82: 2008 - University of Newcastle

2008 ANNUAL REPORT 81

f. Tuyl, R. Gerlach and K.L. Mengersen “Inference for proportions in a 2x2 contingency table: HPD or not HPD?”. Accepted for publication in Biometrics.

f. Tuyl, R. Gerlach and K.L. Mengersen “The rule of three, its variants and extensions”. Accepted for publication in International Statistical Review.

J.C.W. Rayner, D.J. Best and O. Thas* (Eds). “Modern goodness of fit methods”. Accepted for publication in Special Issue of the Journal of Statistical Theory and Practice. Greensboro, North Carolina: Grace Scientific Publishing.

*N. Xiao, L. Xie and M. fu “Kalman filtering over unreliable communication networks with bounded Markovian packet dropouts”. Accepted for publication in International Journal of Robust and Nonlinear Control.

J. Zheng and M. fu “A reset state estimator using an accelerometer for enhanced motion control with sensor quantisation”. Accepted for publication in IEEE Transactions on Control Systems Technology.

CONfERENCE PAPERS

G.J. Adams, B.J. Burke, G.C. Goodwin, J.T. Gravdahl*, R.D. Peirce and A.J. Rojas, “Managing steam and concentration disturbances in multi-effect evaporators via nonlinear modeling and control”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

S. Aphale, B. Bhikkaji and S.O.R. Moheimani “A closed-loop approach to reducing scan errors in nanopositioning platforms”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

S. Aphale, S. Devasia* and S.O.R. Moheimani “Achieving high-bandwidth nanopositioning in presence of plant uncertainties”, Proc. IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Xi’an, China. July 2008.

J.C. Agüero and G.C. Goodwin “Identifiability of EIV dynamic systems with non-stationary data”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

J.C. Agüero, G.C. Goodwin and P.M.J. Van den Hof* “Virtual closed loop identification: A generalized tool for identification in closed loop”, Proc 47th IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December, 2008.

#M. Alamir*, J.S. Welsh and G.C. Goodwin “On useful redundancy in experiment design for nonlinear system identification”, Proc. 47th IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December, 2008.

D. Allingham “Adventures in parallel statistical programming”, Joint Meeting of the 4th World Conference of the International Association for Statistical Computing (IASC), Japan, 5-8 December 2008.

D. Allingham, R.A.R. King, D. Nur and K.L. Mengersen, “further DNA segmentation analysis using approximate Bayesian computation”, Joint Meeting of the 4th World Conference of the International Association for Statistical Computing (IASC), Japan, 5-8 December 2008.

D.J. Best “Comparison of some tests of fit for the Beta-Binomial distribution”, Joint Statistical Meetings (ASA), Denver Colorado, 3-7 August 2008.

B. Bhikkaji, S.O.R. Moheimani and I.R. Petersen “Multivariable integral control of resonant structures”, Proc. IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December 2008

M. Boerlage*, R.H. Middleton and M. Steinbuch* “Rejection of fixed direction disturbances in multivariable electromechanical motion systems”, Proc.17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

S. Chalup, N. Henderson, M. Ostwald and L. Wiklendt “A method for cityscape analysis by determining the fractal dimension of its skyline”, Annual Conference of the Australian and New Zealand Architectural Science Association, Newcastle, 2008.

D. Coutinho*, M. fu and C.E. de Souza* “Output feedback control of linear systems with input and output quantisation”, Proc. IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December 2008.

J.A. De Doná, f. Suryawan, M.M. Seron, and J. Levine* “A flatness-based iterative method for reference trajectory generation in constrained NMPC”, International Workshop on Assessment and future Directions of Nonlinear Model Predictive Control, Pavia, Ital, 5-9 September 2008.

#M.S. Derpich, J. Ostergaard and G.C. Goodwin “The quadratic Gaussian rate-distortion function for source uncorrelated distortions” , Proc. Data Compression Conference (DCC), Utah, USA, 25-27 March 2008.

#M.S. Derpich, D.E. Quevedo and G.C. Goodwin “Conditions for optimality of scalar feedback quantisation”, Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Las Vegas, USA, 30 March-4 April 2008.

A.J. fleming, S. Aphale and S.O.R. Moheimani “A second-order controller for resonance damping and tracking control of nanopositioning systems”, Proc. International Conference on Adaptive Structures and Technologies, Zurich Switzerland, August 2008

J.S. freudenberg*, R.H. Middleton and J.H. Braslavsky “Minimum variance control over a gaussian communication channel”, Proc. 2008 American Control Conference, pp.2625-2630, Seattle, USA, June 2008.

M. fu “Quantisation for feedback control and estimation”, Proc. Chinese Control Conference, Kunming, China, June, 2008.

Page 83: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control82

M. fu, L. Xie* and W. Su* “Connections between quantised feedback control and quantised estimation”, Proc. Int. Conf. on Control, Automation, Robotics and Vision, Hanoi, December, 2008.

M. fu and C.E. de Souza* “State estimation using quantised measurements”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

P.J. Gawthrop*, B. Bhikkaji and S.O.R. Moheimani “Physical-model-based control of a piezoelectric tube scanner”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

B.I. Godoy, J.H. Braslavsky and J.C. Aguero “A simulation study on model predictive control and extremum seeking control for heap bioleaching processes”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

G.C. Goodwin, J.I. Yuz* and J.C. Agüero “Relative error issues in sampled data models”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

H. Haimovich*, E. Kofman* and M.M. Seron “Analysis and improvements of a systematic componentwise ultimate-bound computation method”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

N. Henderson, R. King and S. Chalup “An automated colour calibration system using multivariate gaussian mixtures to segment HSI colour space”, Australasian Conference on Robotics and Automation, Canberra, December 2008.

S. Hovland*, C. Løvaas, J.T. Gravdahl* and G.C. Goodwin “Stability of model predictive control based on reduced-order models”, Proc. 47th IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December, 2008.

E. Kofman*, f. fontenla*, H. Haimovich* and M.M. Seron “Control design with guaranteed ultimate bound for feedback linearizable system”, 17th Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

J. Kulk and J.S. Welsh “A low power walk for the NAO robot”, Australasian Conference on Robotics and Automation, Canberra, December 2008.

P. Lane, R. King and E. Stojanovski “A comparison of Ordinary Least Squares Regression, Hierarchical Regression and Spatial Hierarchical Regression to identify disadvantaged areas in New South Wales”, Australian Statistics Conference (ASC2008). Melbourne, 30 June-3 July 2008.

K. Lau, J.H. Braslavsky, J.C. Agüero and G.C. Goodwin “Application of non-stationary EIV methods to transient electromagnetic mineral exploration”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

Gang Li* and B. Sims “t-Demiclodness principle and asymptotic behaviour for semigroups of nonexpansive mappings in metric spaces”, Proc. 8th International Conference on fixed point theory and its applications, Yokohama Publishers, 2008, pp.103-108.

C. Lovaas, M.M. Seron and G.C. Goodwin “Robust output-feedback MPC with soft state constraints”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

I.A. Mahmood, K. Liu and S.O.R. Moheimani “Two sensor based H-infinity control of a piezoelectric tube scanner”, Proc. 17th IfAC World Congress, Seoul South Korea, 6-11 July 2008

D. Marelli, K. Mahata and M. fu “Linear LMS compensation for timing mismatch in time-interleaved ADCs”, Proc. 34th Annual Conference of the IEEE Industrial Electronics Society (IECON), florida, November 2008.

#J.-J. Martinez*, M.M. Seron and J.A. De Dona. “fault-tolerant switching scheme with multiple sensor-controller pairs”, Proc. 17th IfAC World Congress, Seoul South Korea, 6-11 July 2008

R.H. Middleton and G.J. Adams “Modification of model predictive control to reduce cross-coupling”, Proc. IfAC World Congress, Seoul South Korea, 6-11 July 2008.

S. Mitchell, J.S. Welsh and B. Phung “Relating a distribution transformer’s connection topology and the influence of inductive disparity to the observed frequency response”, Australasian Universities Power Engineering Conference, AUPEC, Sydney, Australia, December, 2008.

D. Nur, K.L. Mengersen and Yan-Xia Lin “Hidden Markov model for DNA sequences segmentation modeling : Simulation and Evaluation”, 24th International Biometric Conference, Dublin, Ireland, 13-18 July 2008.

D. Nur and Yan-Xia Lin “On adaptive estimation in Smooth Threshold Autoregressive (1) models with GARCH(1,1) errors”, Joint Meeting of 4th World Conference of the International Association for Statistical Computing, (IASC), Japan, 5-8 December 2008.

C. Ocampo-Martinez, J. De Doná and M.M. Seron “Actuator fault tolerant control based on invariant set separation”, Proc. 17th IfAC World Congress, Seoul South Korea, 6-11 July 2008

#S. Olaru*, J.A. De Doná and M.M. Seron “Positive invariant sets for fault tolerant multisensor control schemes”, Proc. 17th IfAC World Congress, Seoul South Korea, 6-11 July 2008

#S. Olaru*, J.A. De Doná and M.M. Seron “Receding horizon optimization for control and reconfiguration of multisensor schemes”, International Workshop on Assessment and future Directions of Nonlinear Model Predictive Control, Pavia Italy, 5-9 September 2008

S. Outram and E. Stojanovski “Challenges in teaching skills for medical consultations with year three medical students”, The Association for Health Professional Education (ANZAME). Sydney, 10-13 July 2008.

D.A. Oyarzún*, B.P. Ingalls*, R. H. Middleton and D. Kalamatianos* “Optimal metabolic pathway activation”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

Page 84: 2008 - University of Newcastle

2008 ANNUAL REPORT 83

E. Pereira*, S.O.R. Moheimani and S. Aphale “Op-amp based analog implementation of the integral resonant control scheme”, Proc. International Conference on Adaptive Structures and Technologies, Zurich Switzerland, August 2008.

T. Perez and P. Steinmann “Advanced in gyro-stabilisation of ship roll motion”, PACIfIC 2008 International Pacific Maritime Conference. Sydney, Australia. 29-31 January 2008.

#D.E. Quevedo, A. Ahlén* and G.C. Goodwin “Predictive power control of wireless sensor networks for closed loop control”, Proc. International Workshop on Assessment and future Directions of NMPC, Pavia, Italy, 5-9 September, 2008.

C.R. Rojas, M. Barenthin*, J.S. Welsh and H. Hjalmarsson* “The cost of complexity in identification of fIR systems”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

A.J. Rojas and J.I. Yuz* “Repeated poles in feedback over a class of signal-to-noise ratio constrained channels”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

A.J. Rojas, R.H. Middleton and J.S. freudenberg* “Infimal feedback capacity for a class of additive coloured Gaussian noise channels”, Proc 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008

A.J. Rojas, R.H. Middleton, J.S. freudenberg* and J.H. Braslavsky “Input disturbance rejection in channel signal-to-noise ratio constrained feedback control”, Proc. 2008 American Control Conference, Seattle, Washington, USA, June 2008.

A. Sebastian, A. Pantazi, S.O.R. Moheimani, H. Pozidis and E. Eleftheriou* “A self servo writing scheme for a MEMS storage device with sub-nanometer precision”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

#M.M. Seron, M.E. Romero* and J. A. De Dona “Sensor fault tolerant control of induction motors”, Proc. 17th IfAC World Congress, Seoul South Korea, 6-11 July 2008

#E.I. Silva, D.E. Quevedo and G.C. Goodwin “Optimal controller design for networked control systems”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

#E.I. Silva, G.C. Goodwin and D.E. Quevedo “On networked control architectures for MIMO plants”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

S. Snodgrass, D. Rivett, V. Robertson and E. Stojanovski “Accuracy of applied forces during cervical mobilisation improves with real-time objective feedback.”, International federation of Orthopaedic Manipulative Therapists (IfOMT) Rotterdam, The Netherlands, 8-13 June 2008.

S. Snodgrass, D. Rivett, V. Robertson and E. Stojanovski “Measurement of manual forces applied by physiotherapy students learning cervical spine mobilisation skills”, International federation of Orthopaedic Manipulative Therapists (IfOMT), Rotterdam, The Netherlands, 8-13 June 2008.

E. Stojanovski, P. Lane and R. King “Spatial Assessment of Crime”, 4th World Conference of the International Association for Statistical Computing (IASC),Yokohama, Japan, 5-8 December 2008.

E. Stojanovski and K. Mengersen “Measuring uncertainty in a predictive model: A case study”, 24th International Biometric Conference (IBC 2008), Dublin Ireland, 13-18 July 2008.

B. Stokes “The entropy concentration theorem – Explanation and application”, International Mathematica Symposium, Maastricht, The Netherlands, 20-24 June 2008.

B. Stokes “Synthetic survival data generation from Kaplan-Meier Curves”, International Mathematica Symposium, Maastricht, The Netherlands, 20-24 June 2008.

#M. Wang, G.C. Goodwin and D.E. Quevedo “EM based receiver design for uplink MIMO-OfDMA Systems”, Proc. IEEE International Conference on Communications (ICC), Beijing, China, 19-23 May, 2008.

#M. Wang, D.E. Quevedo, G.C. Goodwin and B.S. Krongold “A complex-baseband active-set approach for tone reservation PAR reduction in OfDM systems”, Proc. Australian Communications Theory Workshop (AusCTW), Christchurch, New Zealand, 30 Jan-1 feb, 2008.

J.S. Welsh, T. Daredia, f. Sobora, L. Vlacic and G.C. Goodwin “Simulated versus hardware laboratories for control education: A critical appraisal”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

V. Wertz*, E.I. Silva, G.C. Goodwin and B. Codrons* “Performance limitations arising in the control of power plants”, Proc. 17th IfAC World Congress, Seoul, South Korea, 6-11 July, 2008.

A. Wong and S. Chalup “Towards visualisation of sound-scapes through dimensionality reduction”, 2008 IEEE World Congress on Computational Intelligence (WCCI), Hong Kong, June 1-6, 2008.

A. Wong and S. Chalup “Sound-scapes for robot localisation through dimensionality reduction”, Australasian Conference on Robotics and Automation, Canberra, December 2008.

Y.K. Yong, S. Aphale, S.O.R. Moheimani “Design, analysis and control of a fast nanopositioning stage”, Proc. IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Xi’an, China, 2-5 July, 2008.

Y.K. Yong, K. Liu and S.O.R. Moheimani “H-infinity control for reducing cross-coupling in a compliant XY nanopositioning stage”, Proc. International Conference on Adaptive Structures and Technologies, Zurich Switzerland, August 2008.

J. Zheng and M. fu “A reset state estimator for linear systems to suppress sensor quantisation effects”, Proc. 17th IfAC World Congress, Seoul, South Korea, June, 6-11 July 2008.

I.A. Mahmood, K. Liu and S.O.R. Moheimani “Two sensor based H-Infinity control of a piezoelectric tube scanner”, Proc. 17th IfAC World Congress Seoul, South Korea, 6-11 July 2008.

Page 85: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control84

TECHNICAL REPORTS TO INDUSTRY:

BHP Billiton Innovation

BHP B/IMP/08/01M. Zhang“An Automated Practical Mining Phase Design Using Simulated Annealing”1 April 2008

BHP B/IMP/08/02T. Perez and G.C. Goodwin“Ideas on Identification of Commodity Price Models”15 April 2008

BHP B/IMP/08/03T. Perez, G.C. Goodwin, K. Barbosa* and J.-C. Aguero “An Algorithm for Obtaining Initial Parameter Estimates of Two-factor Commodity Price Models” 8 May 2008

BHP B/IMP/08/04 G.C. Goodwin, T. Perez and J.-C. Aguero “Some Brief Observations on Empirical Mode Decomposition” 7 August 2008

BHP B/IMP/08/05 M. fu and Xin Tai* “On Parameter Estimation of the Schwartz-Smith Two-factor Model” 10 July 2008

BHP B/IMP/08/06 T. Perez, G.C. Goodwin and B. Godoy“A Prediction-Error Approach to the Identification of Two-factor Commodity Price Models” 26 September 2008

BHP B/IMP/08/07 Xin Tai* and M. fu“Parameter Estimation of the Schwartz-Smith Two-factor Model for Aluminium and Copper Data of BHP-Billiton” 30 September 2008

BHP B/IMP/08/08T. Perez, G.C. Goodwin and B. Godoy“A Two-Step Approach to Parameter Estimation of Commodity Price Models”19 December 2008

BHP B/OBOG/08/1K. Lau and J.H. Braslavsky“Sferics Project Report 19: X-Y to Z Model Estimation using Auxiliary XY Measurements (Mount Keith Data, Preliminary Results)”15 January 2008

BHP B/OBOG/08/02K. Lau and J.H. Braslavsky“Sferics Project Report 20: X-Y to Z Model Estimation using Auxiliary XY Measurements (Mount Keith Data, further Results)” 25 february 2008

BHP B/OBOG/08/03K. Lau and J.H. Braslavsky“Sferics Project Report 21: New Sensor (HBI) Response Compensation”31 March 2008

BHP B/OBOG/08/04 K. Lau, D. Ugryumova* and J.H. Braslavsky“Sferics Project Report 22: X-Y to Z Transfer function Estimation, Sferics Cancellation on Transmitter-on Data, and Magnetotellurics Impedance Estimation” 26 May 2008

BHP B/OBOG/08/05 K. Lau and J.H. Braslavsky“Sferics Project Report 23: Matlab Implementation of frequency Response Estimation and Noise Cancellation Algorithms”28 July 2008

BHP B/OBOG/08/06K. Lau and J.H. Braslavsky“Sferics Project Report 24: frequency Response Estimation Using Transmitter-on Data (March 2008 Data)”18 August 2008

BHP B/OBOG/08/07D. Ugryumova* “System Identification and Noise Cancellation in Electromagnetic Mineral Exploration” 29 September 2008

BHP B/OBOG/08/08K. Lau and J.H. Braslavsky“Sferics Project Report 25: Preliminary Data Analaysis – Total field Sensors (Sept. 2008 Data)” 20 October 2008

BHP B/OBOG/08/09K. Lau and J.H. Braslavsky“Sferics Project Report 25: Low frequency Noise Analysis – Total field Sensors (Sept. 2008 Data)” 3 December 2008

BHP B/Cogen/08/01G.J. Adams, G.C. Goodwin, J.T. Gravdahl* and A. Rojas“BHP Co-generation Project Description” 9 April 2008

Connell Wagner

CW/SYNC/08/01D. Allingham and J.S. Welsh“Report 1: Preliminary Synchronous Machine Parameter Estimation” 1 October 2008

CW/SYNC/08/02 D. Allingham and J.S. Welsh“Report 2: Synchronous Machine Parameter Estimation: Bayswater Experimental Data”31 October 2008

CW/SYNC/08/03 D. Allingham and J.S. Welsh“Report 3: Preliminary q-axis Results” 6 November 2008

CSR

CSR/Brake/08/01A. Rayner “CSR Brake Van Control” 8 february 2008

CSR/Evap/08/01 G.J. Adams“Analysis of Valve Tests” 29 September 2008

CSR/Evap/08/2G.J. Adams and A.J. Rojas“CSR Evaporator Control – Brix Control Improvements”19 December 2008.

Page 86: 2008 - University of Newcastle

2008 ANNUAL REPORT 85

Matrikon

Mat/ADCC/08/1A.M. Medioli “Next generation model-based control tools – Automated Downtime Cause Classifier Report 1”15 September 2008.

Mat/ADCC/08/2A.M. Medioli“Next generation model-based control tools – Automated Downtime Cause Classifier Report 2”19 December 2008.

Mat/NGMT/08/01 G.J. Adams “Progress Report for february 2008”28 february 2008

Mat/NGMT/08/02 G.J. Adams “Progress Report for May 2008”22 May 2008

Mat/NGMT/08/03 G.J. Adams“Next Generation Model-Based Control Tools – Implementation of Decoupling in CPOmpc”27 June 2008

Mat/NGMT/08/04G.J. Adams“State space models for Cpompc” 18 August 2008

Mat/NGMT/08/05 G.J. Adams“Progress report for August 2008” 19 August 2008

Mat/NGMT/08/06 G.J. Adams“Next Generation Model-Based Control Tools – CPOmpcid Review”9 September 2008

Mat/NGMT/08/07 N. Germyn“Next Generation Model-Based Control Tools – Steady State Optimisation Tools”29 October 2008

Mat/NGMT/08/08 G.J. Adams and G.C. Goodwin“Next Generation Model-Based Control Tools – Explicit Non-linear MPC for CPO”30 October 2008

Mat/NGMT/08/09 G.J. Adams“Next Generation Model-Based Control Tools – Progress Report for November 2008”20 November 2008

Robotiker-Tecnalia

Robotiker/08/01 T. Perez“Specifications for the Implementation of a Simulation Tool for a Gyro Wave Energy Converter” 10 March 2008

Robotiker/08/02 T. Perez“Performance Analysis of a Gyro-based Wave Energy Converter Under Linear Control” 30 April 2008

Halcyon International

Halcyon/08/01 T. Perez“Analysis and Design of a Gyro-stabiliser Linear Damping Controller” 4 March 2008

Halcyon/08/02 T. Perez“Gyro-stabiliser Precession Damping Automatic Grain Controller (AGC)” 12 March 2008

Hamilton Jet:

CFW-HJ/Exp-Design/08/01 T. Perez and G.C. GoodwinCommercial-in-Confidence Report25 february 2008

CFW-HJ/Exp-Design/08/02 T. Perez and C. Lovaas Commercial-in-Confidence Report7 May 2008

CFW-HJ/Exp-Design/08/03C. Lovaas and T. PerezCommercial-in-Confidence Report29 May 2008

Offshore Simulation Centre (Norway)

OSC/FD-Ident/08/01T. Perez“Use and Implementation of a frequency-Domain Identification Method for Seakeeping Models of Marine Structures” 12 May 2008

University of Newcastle, School of Biomedical Sciences

DSC/BME/08/01A.J. Rojas“Analysis of Membrane Currents in the Time and frequency Domains” 27 December 2008

Page 87: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control86

PERfORMANCE INDICATORS REPORTPERfORMANCE INDICATORS REPORT

08

Page 88: 2008 - University of Newcastle

2008 ANNUAL REPORT 87

RESEARCH fINDINGS AND COMPETITIVENESS

Description 2008 Actual Details of 2008 outcomes 2008 -2010 Target

Refereed International Journal J: 57 See Publications, pages 77-83. 150 & Conference Publications C: 59

Number of Patents 1 See Publications, page 77. 2

Invitations to address & participate in 17 Agüero, Braslavsky, Goodwin and Lau 15 international conferences Invited Session: J.-C. Agüero, G.C. Goodwin and P.M.J. Van den Hof, “Virtual closed loop identification: A generalised tool for identification in closed loop”, Proc. 47th IEEE Conference on Decision and Control, Cancun, Mexico, 9-11 December 2008

Invited Session: J.-C. Agüero and G.C. Goodwin, “Identifiability of EIV dynamic systems with non-stationary data”, 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

Invited Session: K. Lau, J.H. Braslavsky, J.-C. Agüero and G.C. Goodwin, “Application of non-stationary EIV methods to transient electromagnetic mineral exploration”, 17th IfAC World Congress, Seoul, South Korea, 6-11 July 2008.

Fleming – Invited Session: J. Maess, A.J. fleming and f. Allgöwer, “Model-based vibration suppression in piezoelectric tube scanners through induced voltage feedback”, Proc. American Control Conference, Seattle, WA, June, 2008.

Fu – Invited panelist in Plenary Panel Session: Control of Complex Systems, Int. Conf. on Control, Automation, Robotics and Vision, Hanoi, December, 2008.

– Invited panelist in Plenary Panel Session at Chinese Control Conference, Kunming, July 2008.

– Invited session: M. fu, L. Xie and W. Su, “Connections between quantised feedback control and quantised estimation,” Int. Conf. on Control, Automation, Robotics and Vision, Hanoi, December, 2008.

Mengersen – Program Chair, International Society for Bayesian Analysis (ISBA) World Meeting, Hamilton Island, 21-25 July 2008.

Also, see Publications, pages 77-84.

Page 89: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control88

Invitations to visit international laboratories 19 Agüero – Universidad Santa Maria, Chile. 15

Fu – Nanyang Technological University, Singapore; Zhejiang University, China; Shandong University, China; Southern China University of Science and Technology, China.

Goodwin – University Uppsala, Sweden.

Marelli – Acoustics Research Institute, Austrian Academy of Science, Austria; CNRS, Grenoble, france.

Mengersen – NCEAS, Santa Barbara, USA; Dauphine University, Paris, france; South African Statistical Society and National Cheetah Conservation Groups.

Moheimani – Automatic Control Labs, ETH, Switzerland.

Perez – NTNU Norwegian University of Science and Technology, Norway.

Rayner – Texas A&M, College Station, USA; Ghent University, Belgium.

Seron – NTNU Norwegian University of Science and Technology, Norway; Grenoble Institute of Technology, france.

Zheng – Southern China University of Science and Technology, China; NTU, Singapore.

Number of commentaries about the 4 See Industry Interaction and 6 Centre’s achievements Selected Outcomes, page 12.

Additional competitive grant income $170,000 ARC Discovery Grants. $150,000

RESEARCH TRAINING AND PROfESSIONAL EDUCATION

Description 2008 Actual Details of 2008 outcomes 2008 -2010 Target

Postgraduate students recruited 3 Brendan Burke, Aurelio Salton, 18 Alain Yetendje-Lemegni. (See page 7).

Postgraduate completions 6 Research Higher Degrees awarded in 2008: 15 Dr Boris Godoy, Dr Christian Løvaas, Dr Jose Mare, Dr Adrian Medioli, Dr Trevor Moffiet, Dr Cristian Rojas. (See page 8).

Supervise Honours students 10 Matthew fairbairn, “Control and optimisation 24 of a linear inertial drive” (fleming)

Megan ford “A comparison of Beta-Binomial Posterior-Predictive and CUSUM control charts for clinical indicator data” (Howley)

Nicolas Germyn “Steady state optimisation tools” (Adams)

Peter McPherson, “Dynamically allocating plant sequencer and communication system” (fu)

Antony Mujic, “force feedback in teleoperated surgery” (Welsh)

PERfORMANCE INDICATORS REPORT

Page 90: 2008 - University of Newcastle

2008 ANNUAL REPORT 89

Andrew Paver, “Control of a 2 wheeled robot” (Welsh)

Tristan Rayner, “Control of a 2 wheeled robot” (Welsh)

Ian Robinson “Determining the cost of smoking to the Hunter Area Health Service” (Stojanovski)

Aaron Thompson – “Dynamic allocation by conductor sag modelling” (fu)

Ben Tillman, “Iterated monodromy groups” (Willis)

Professional courses run 6 See Conferences, Courses and Workshops, 3 page 14.

Participation in professional courses 0 15

Number and level of undergraduate and 6 n ELEC2320 “Electrical Circuits” (Welsh) 15 high school courses in the priority areas n ELEC2400 “Signals & Systems” (fu) undergraduate n ELEC3400 “Signal Processing” (De Doná) courses n ELEC3850 “Electrical Engineering Design” (Welsh) n ELEC4400 “Automatic Control” (fleming) n ELEC4410 Control System Design Management (Adams, Braslavsky, Rojas and Welsh)

INTERNATIONAL, NATIONAL AND REGIONAL LINKS AND NETWORKS

Description 2008 Actual Details of 2008 outcomes 2008 -2010 Target

Published papers with international J: 30 See Publications, pages 77-83. 60 co-authors C: 20

International Visitors 24 See Visitors, pages 10-11. 30

Collaborative national and international 6 See pages 14 and 17. 15 workshops and exchanges

Visits to overseas laboratories 20 Agüero – Universidad Santa Maria, Chile. 45

Fu – Nanyang Technological University, Singapore; Zhejiang University, China; Shandong University, China; Southern China University of Science and Technology, China.

Goodwin – University of Uppsala, Sweden.

Marelli – Acoustics Research Institute, Austrian Academy of Science, Austria; CNRS, Grenoble, france.

Mengersen – NCEAS, Santa Barbara, USA; Dauphine University, Paris, france; South African Statistical Society and National Cheetah Conservation Groups.

Moheimani – IBM Research Labs, Zurich, Switzerland; Automatic Control Lab, ETH, Switzerland.

Page 91: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control90

Perez – NTNU Norwegian University of Science and Technology, Norway.

Rayner – Texas A&M, College Station, USA; Ghent University, Belgium.

Seron – NTNU Norwegian University of Science and Technology, Norway; Grenoble Institute of Technology, france.

Zheng – Southern China University of Science and Technology, China; NTU, Singapore.

Memberships of national and international 14 Braslavsky – Associated Editor IET 15 professional committees Control Theory and applications (UK)

Welsh – Associate Editor IEEE Control Systems Society Conference Editorial Board; Member IfAC Technical Committee on Mechatronic Systems; IPC Member IfAC2010 Symposium on Mechatronic Systems; Member HSC Examination Committee for Engineering Studies (NSW).

Goodwin – International Advisory Committee of 2008 Control and Decision Conference in Yantai, China, 2 July 2008

Moheimani – Vice-Chair IfAC Technical Committee on Mechatronics Systems; Associate Editor IEEE Transactions on Control Systems Technology; Associate Editor International Journal of Control, Automation and Systems; IPC Member, Australian Robotics and Automation Conference; IPC Member SPIE 15th International Symposium on Smart Structures and Materials; IPC Member, 2008 IfAC World Congress, Seoul, Korea; Member IfAC Technical Committee on Robust Control.

Fleming – Associate Editor Advances in Vibration and Acoustics.

Research projects with 21 See Research Programs section. 15 international partners

END-USER LINKS

Description 2008 Actual Details of 2008 outcomes 2008 -2010 Target

Commercialisation activities 2 Virtual Laboratories (Goodwin); 3 Positioning System and Method – Patent filed (fleming)

Government, industry and business briefings 7 Boeing Research and Technology, Australia; 6 Defence Materials Organisation; Energy Australia; fusion Group, Australian National University; Hydro Aluminium; John Hunter Hospital; Rio Tinto.

PERfORMANCE INDICATORS REPORT

Page 92: 2008 - University of Newcastle

2008 ANNUAL REPORT 91

Centre associates trained in technology 3 Adams, fleming, Moheimani. 3 transfer and commercialization

Public Awareness programs 0 3 national broadcasts

Cash contributions from end-users to the $215,000 Boeing Research and Technology Australia, $50,000 Centre, including research contracts BHP-Billiton, Connell Wagner, Hatch IAS, Matrikon.

In-kind contributions from end-users to $354,420 $100,000 the Centre

ORGANISATIONAL SUPPORT

Description 2008 Actual Details of 2008 outcomes 2008 -2010 Target

Annual cash contributions from $633,142 BHP-Billiton $100,000; Matrikon $50,000; $500,000 pa collaborating organisations The University of Newcastle $300,000; NSW Department of State and Regional Development $161,542; Queensland University of Technology $21,600

Annual in-kind contributions from $3,402,464 $1M pa collaborating organisations

New organisations recruited to or 3 Boeing Research and Technology, Australia; 3 involved in the Centre CfW Hamilton Jet, New Zealand; Norwegian University of Science and Technology.

Annual cash contributions from $65,000 Boeing Research and Technology, Australia; $20,000 pa other organisations Connell Wagner; Hatch IAS.

Annual in-kind contributions from $43,000 $40,000 pa other organisations

GOVERNANCE

Advisory Board 1 See Advisory Board, page 9. 3

Strategic Plan endorsed by Advisory Board 1 Accepted on 25 July 2008. 1

Management of Centre nodes 6 Centre Leaders and Deputies Meetings. 6 pa

Centres Key performance measures 1 2008 Annual Report. 3

NATIONAL BENEfIT

Measures of expansion of Australia’s N/A See Industry Interaction and N/A capability in the priority areas Selected Outcomes, page 12.

Case studies of economic, social, cultural 2 n feasibility study of new electromagnetic 3 sensors or environmental benefits for mineral exploration. (See project A.2 page 25). n financial and commodity modelling. (See project A.3 page 28).

Page 93: 2008 - University of Newcastle

ARC Centre of Excellence for Complex Dynamic Systems and Control92

Acc

ount

nam

e

Cen

tral

Acc

ount

s

Con

trol

Sys

tem

S

igna

l in

dust

rial

Mec

hatr

onic

s D

istr

ibut

ed

Mat

hem

atic

al

Bay

sian

To

TAl

(C

ombi

ned)

D

esig

n P

roce

ssin

g C

ontr

ol

S

ensi

ng a

nd

Sys

tem

s le

arni

ng

& o

ptim

isat

ion

C

ontr

ol

Theo

ry

(Uon

)

(C

ombi

ned)

Acc

oun

t N

umb

ers

10.8

0428

; 804

46;

10.8

0429

10

.804

30

10.8

0431

; 10

.804

32

$10

.815

69

$10

.813

82

$10

.813

83

80

456;

804

58;

8044

3; 8

0444

;

8043

5; 3

0777

80

442,

804

80;

8168

8; 8

1568

;

81

786

Inco

me

20

07 b

/f

647,9

11

(20,

964)

19

3,48

3 24

8,33

0 52

,167

-

126,

508

115,

501

1,36

2,93

6 A

RC

08

gran

t 16

0,17

8 48

0,00

0 46

,000

63

0,00

0 34

5,00

0 75

,000

57

,000

80

,000

1,

873,

178

Oth

er In

com

e 16

0,00

0 -

-

55

,000

-

-

-

-

21

5,00

0 Th

e U

nive

rsity

of N

ewca

stle

Sup

port

26

2,50

0 -

-

- -

- -

37,5

00

300,

000

NS

W D

ept S

tate

& R

egio

nal D

ev

161,

542

- -

- -

- -

- 16

1,54

2 R

educ

tion

of 2

007

b/f

(55,

313)

-

- -

- -

- -

(55,

313)

(D

ue to

fE

BE

’s ti

ght b

udge

t)

Tota

l Inc

om

e 1,

336,

819

459,

036

239,

483

933,

330

397,1

67

75,0

00

183,

508

233,

001

3,85

7,34

4

Sal

ary

Exp

endi

ture

Sal

arie

s (A

cade

mic

) 24

2,09

1 26

2,27

2 11

7,545

42

7,144

15

8,79

3 9,

691

1,142

54

,179

1,27

2,85

7 S

alar

ies

(Gen

eral

) 11

7,346

-

-

4,86

9 7,4

38

- -

29,7

58

159,

412

Sal

arie

s on

-cos

ts

114,

695

92,

258

39,5

33

156,

973

52,3

61

3,28

7 3

10

24,3

34

483,

752

Sch

olar

ship

s/S

tude

nt S

uppo

rt

7,960

11

,733

10

,652

4,

930

3,37

3

9,33

6 1,

520

5,00

0 54

,504

Tota

l Sal

ary

& R

elat

ed C

ost

s 48

2,09

1 36

6,26

2 16

7,73

0 59

3,91

7 22

1,96

5

22,3

15

2,97

1 11

3,27

1 1,

970,

524

No

n-S

alar

y E

xpen

ditu

re

C

onsu

mab

les

30,5

72

3,44

8 4,

501

5,47

3 4

,971

1,

210

66

16

2

50,0

78

Ser

vice

s 7

42

-

- -

- -

- -

742

Tr

avel

8,

796

69,0

52

9,15

4 17

,911

24

,908

5,

424

9,51

2 29

,513

17

4,27

0 R

epai

rs &

Mai

nten

ance

-

-

-

-

-

-

-

-

-

Util

ities

3

80

- 1

8

464

7

45

18

-

-

1,

625

Equ

ipm

ent

122,

865

141

2,

409

2,64

4 3,

794

1,08

0 46

,184

3,72

1 18

2,83

8 O

ther

s 5,

772

33,5

26

474

3

63

1,29

6 8

5

-

-

41,5

17

Sha

red

Res

earc

h gr

ant P

aym

ent

180,

000

- -

- -

- -

- 18

0,00

0 V

isito

r 4,

217

13,9

28

9,83

9 -

- -

- -

27,9

84

Tota

l No

n-S

alar

y E

xpen

ditu

re

353,

343

106,

167

30,4

84

36,6

94

35,7

14

7,81

7 55

,762

33

,073

65

9,05

4

Tota

l Exp

endi

ture

s

835,

435

472,

430

198,

215

630,

611

257,

679

30,1

32

58,7

33

146,

344

2,62

9,57

8

Bal

ance

as

at 3

1/12

/200

8 50

1,38

4 (1

3,39

4)

41,2

68

302,

719

139,

489

44,8

68

124,

775

86,6

57

1,22

7,76

6

INC

OM

E A

ND

EX

PE

ND

ITU

RE

STA

TEM

EN

T

fOR

TH

E Y

EA

R E

ND

ED

31S

T D

EC

EM

BE

R 2

008