Hur fr man system att uppfra sig som man vill? Reglerteknik och
intressanta tillmpningar Carl-Fredrik Lindberg, 2013-04-25
Slide 2
Single and Double Inverted Pendulum Movie found on internet
http://vimeo.com/2952236 2
Slide 3
Automatic Control (Reglerteknik in Swedish) 3 Automatic control
gives: Stability Fast disturbance rejection Small variations in
controlled variables Etc. Modeling the tank dynamics Tuning the PID
controller
Slide 4
How a control engineer sees the world 4 Set- points Measurement
noise Measureable disturbances Non measureable disturbances
Controller Process Feedback Feed- forward Control signals Measured
outputs
Slide 5
Automatic Control 5 Modeling Physical models, grey-box,
black-box, Controller design Feedback, feedforward, stability,
performance, robustness, adaptive, Estimation States, model
parameters, Diagnostics Fault detection and isolation,
Optimization
Slide 6
Project examples 6
Slide 7
EM Stabilizer in Galvanizing Lines 7 Stabilization of strip
gives reduced zinc over-coating
Slide 8
EM Stabilizer in Galvanizing Lines 8 Improved strip position
damping with EM Stabilizer
Slide 9
EM Control 9 Simulation of the flow Validation of model with
PIV measurements in water model Continuous casting process Water
model Control of the steel flow in the mold gives higher steel
quality
Slide 10
Paper Machine Steam Energy Fingerprint 10 By quantifying steam
flows and identifying energy waste, suggestions on how to improve
energy efficiency are given
Slide 11
Arc Furnace Control 11 By control, optimization and EMS, the
yield, energy efficiency and productivity is increased
Slide 12
SVC Voltage Control with Feedforward Improved disturbance
rejection after TSC switching 12
Slide 13
Wireless Control New possibilities with wireless control
Increased flexibility to install sensors Cost for wiring disappears
Large challenges Reliability Latency Safety Co-existence with other
wireless systems 90% of the communication could often be saved by
smarter sampling Different sampling strategies have been developed
Sample as seldom as possible without losing control performance
Taking into account that information packages may be delayed or
lost 13
Slide 14
Diagnostics Control loop diagnostics On-line methods which
automatically alarms for poor control loop performance. Examples
Detection of oscillations and changed signal variance. Control
performance measurements, Harris index etc. Classification of some
faults, e.g. valve stiction, external disturbances. 14 Process
section diagnostics Faults are both detected and isolated Fault
examples: Detection of bias in a sensor Detection of clogging of a
valve Detection of erosion in a pump Track heat transfer rate in
heat exchanger Benefits: Early detection gives higher availability
Reduced maintenance costs. Avoids wrong dosing when sensor has
drifted. Example of fault isolation. The red bar indicates the most
probable faulty signal.
Slide 15
Waste Water Treatment Some examples from my Ph.D. 15 Control of
dissolved oxygen concentration Estimation of dissolved oxygen
dynamics Adaptive control of external carbon flow rate
