CONTROL SYSTEM
DESIGN
ELEC ENG
4CL4
Winter 2004
CALENDAR:
Design of linear control systems using classical and computer-aided
techniques; design and performance limitations; sampled-data control;
architectural issues.
Three lectures, one tutorial, one lab (every other week);
second term
Prerequisite: ELEC ENG 3TP4 or ELEC ENG 3CK4
COURSE OBJECTIVES:
To develop a wide range of techniques for the design of control systems and
an appreciation of their practical application. This involves the analysis of
simplified mathematical models for the dynamic system to be controlled, followed
by the synthesis of a controller based not only on this model, but also on
knowledge of the limitations of the model and the control system architecture.
COURSE LOADING:
- Lectures: 3 1-hour lectures per week
- Tutorial: 1 1-hour tutorial per week
- Lab: 1 3-hour session every other week
- Prelab preparation and
reports, homework assignments:
2.5 hrs per week
- Study time: 3 hours per week
- Total hours per week: 11
CEAB WEIGHTING:
TEXTBOOK:
- G. C. Goodwin, S. F. Graebe, and M. E. Salgado, "Control System Design",
Prentice Hall, 2001
RECOMMENDED READING:
- R. C. Dorf and R. H. Bishop, "Modern Control Systems", ninth edition,
Prentice Hall, 2001.
DETAILED COURSE CONTENT:
Introduction to the Principles of Feedback (2 hours)
Modelling (2 hours)
- Introduction to the goals and principles of modelling
- Differential and difference equation models
- State-space models
- Linearization
- Modelling errors
Continuous-Time Signals and Systems (3 hours)
- Review of the Laplace and Fourier transforms
- Evaluation of time and frequency
responses from poles and zeros
- Frequency response of modelling errors
Analysis of Single-Input Single-Output
(SISO) Control Loops (5 hours)
- Sensitivity functions
- Stability analysis
- Root Locus, frequency response
techniques
- Relative stability
- Robustness
Classical Proportional-Integral-Derivative (PID)
Control (3 hours)
- Ziegler-Nichols and Cohen-Coon methods
- Lead-Lag compensators
- Industrial application: Control of a distillation column
Synthesis of SISO Controllers (2 hours)
- Pole assignment
- Smith predictors
Fundamental Design Limitations in
SISO Control (6 hours)
- Limitations caused by sensors and actuators, noise and disturbances, and
modelling errors
- Structural constraints
- Interpolation constraints
-
Time-domain steady-state error and integral constraints
- Remedies
Frequency Domain Design Limitations (3 hours)
- Bode's integral constraints
- Poisson's integral constraints
- Examples of design trade-offs
Models for Sampled-Data Control
Systems (4 hours)
- Introduction to the motivations and design issues for digital control
-
Sampling and reconstruction
- Shift operator and Z-transform
- Delta
operator and discrete Delta-transform
- Discrete models for sampled
continuous systems
- Frequency response of sampled data systems
Digital Control (3 hours)
- Relationship to continuous-time control loop analysis
-
Approximate-continuous digital designs
- At-sample digital designs
-
Internal model principle for digital design
Architectural Issues in SISO Control (3 hours)
- Disturbance rejection and exact reference tracking via the internal model
principle
- Feedforward control
- Cascade control
(Total Course = 36 hours)LABORATORIES:
- Lab 1: MATLAB/Simulink simulation of continuous casting.
- Lab 2:
MATLAB/Simulink analysis of control system stability.
- Lab 3: PID
tuning in MATLAB/Simulink.
- Lab 4: Analysis of control design
limitations and remedies within MATLAB/Simulink.
- Lab 5: Simulation
of digital control within MATLAB/Simulink.
Latest Update July 14, 2003, by Dr. Ian C. Bruce