13
QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ. of Illinois at Urbana-Champaign Notre Dame, Oct 18, 2005

QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

Embed Size (px)

Citation preview

Page 1: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

QUANTIZATION and DELAY EFFECTS

in NONLINEAR CONTROL SYSTEMS

Daniel Liberzon

Coordinated Science Laboratory andDept. of Electrical & Computer Eng.,Univ. of Illinois at Urbana-Champaign

Notre Dame, Oct 18, 2005

Page 2: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

PROBLEM FORMULATION

QUANTIZER DELAY

Architecture-independent approach

Page 3: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

QUANTIZATION ONLY

quantization error

Assume

Page 4: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

QUANTIZATION ONLY

Solutions that start in

enter and remain there

quantization error

Assume

This is input-to-state stability (ISS) w.r.t. actuator errors

In time domain: [Sontag ’89]

Page 5: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

QUANTIZATION and DELAY

Assuming ISS w.r.t. actuator errors as before:

In time domain:

where

Page 6: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

SMALL – GAIN ARGUMENT

hence

ISS property becomes

if

then we recover ISS w.r.t. [Teel, IEEE TAC, Jul ’98]

Small gain:

Page 7: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

FINAL RESULT

small gain true

Need:

Page 8: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

FINAL RESULT

small gain true

Need:

Page 9: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

FINAL RESULT

small gain true

Need:

Page 10: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

FINAL RESULT

Need:

small gain true

Page 11: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

FINAL RESULT

solutions starting in

enter and remain there

Need:

small gain true

Page 12: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

VARIATIONS

Relaxing ISS assumption

Previous approach works if is sufficiently small

Different system architecture

globally asymptotically stabilizing we have

External disturbances

DELAY

QUANTIZER

Need ISS with respect

to measurement errors

(or work with small )

Page 13: QUANTIZATION and DELAY EFFECTS in NONLINEAR CONTROL SYSTEMS Daniel Liberzon Coordinated Science Laboratory and Dept. of Electrical & Computer Eng., Univ

DYNAMIC QUANTIZATION

zoom in

Can recover global asymptotic stability

– zooming variable

Hybrid quantized control: is discrete state

zoom out

ISS from to

[Nesic-Liberzon ’05]

ISS from to small-gain condition