Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
279 views

ECE.488 Multi-Various Analog Controls

This document provides information about the ECE488 - Multivariable Control Systems course. The course objectives are to help students think like advanced control engineers and become familiar with powerful control analysis and design tools. Students will learn about feedforward and feedback control, performance limitations, robustness analysis, and multi-input multi-output systems. Main applications studied include mass-spring-damper systems, control of rockets and airplanes, and a demo helicopter system. The course involves lectures, tutorials, project exercises, midterm exams, and a final exam. Students can work in groups on the project exercises, which apply techniques from lectures to control systems.

Uploaded by

phoebezz
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
279 views

ECE.488 Multi-Various Analog Controls

This document provides information about the ECE488 - Multivariable Control Systems course. The course objectives are to help students think like advanced control engineers and become familiar with powerful control analysis and design tools. Students will learn about feedforward and feedback control, performance limitations, robustness analysis, and multi-input multi-output systems. Main applications studied include mass-spring-damper systems, control of rockets and airplanes, and a demo helicopter system. The course involves lectures, tutorials, project exercises, midterm exams, and a final exam. Students can work in groups on the project exercises, which apply techniques from lectures to control systems.

Uploaded by

phoebezz
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

ECE488 Multivariable Control Systems

Instructor

Prof D.E. Davison, ddavison@uwaterloo.ca

Required background

Any first course in control (ECE 380, MTE 360, SYDE 352, ), and any course in linear
algebra.

Course objectives

The two high-level course objectives are:


1. Learn to think more like an advanced control engineer.
2. Become familiar with more powerful control analysis and design tools.
Lower-level objectives include:
(a) Recognize the benefits of feedforward and feedback control schemes.
(b) Explain fundamental performance limitations that arise in feedback control
because of pole or zero locations in the plant.
(c) Analyze a control system to determine how robust it is with respect to uncertainty
in the plant model.
(d) Understand complexities that arise in a multi-input multi-output (MIMO) setting,
in particular the issue of cross-channel interaction.
(e) Perform basic calculations on a MIMO system (e.g., find the poles and zeros,
interpret a MIMO Bode plot).
(f) Become familiar with the state-space framework as an alternative to the transfer
function framework, and recognize the benefits of the state-space approach,
especially for MIMO systems.
(g) Be proficient at manipulating MIMO block diagrams using both transfer function
methods and state-space methods.
(h) Design controllers for MIMO systems using various transfer function techniques.
Recognize situations where such techniques work well and do not necessarily
work well.
(i) Master several standard, and very powerful, state-space control design techniques
for MIMO systems.
(j) Be able to explain how state-space methods and transfer-function methods
complement one another.
Main applications studied in class:

simple mass-spring-damper systems


control of a rocket (a tricky balancing problem!)
control of a 747 airplane (a true multivariable system)
a demo helicopter system (tethered for safety)

1 of 3

ECE488 Practical Details and Project Exercises


Textbook

An extensive set of lecture notes, exercises and solutions, plus all old exams will be
available for purchase. Other references will be available for those who request them.
(No textbook needs to be purchased.)

Scheduling for W2013 Lectures on Mondays, Wednesdays, and Friday 1:30-2:20pm


(tentative)
Tutorials on Wednesdays 9:00-9:50pm
Marking scheme

Project exercises
Midterm exams (2 short ones)
Final exam

Project exercises

There is a set of six project exercises that you will work on throughout the term.
The exercises are provided at the end of most sections of the notes. Each project
exercise applies the methods of that section to a particular control system. The
purpose of the project exercises is to increase your understanding of the techniques
and ideas presented in the lectures by applying them to a quasi-realistic system. By
working through the problems, you will also increase your familiarity with a
popular numerical software package (Matlab) and learn to appreciate the
limitations of such software. Project exercises can be done in groups, and the
instructor will gladly work with students on any aspect of the exercises. The
purpose of the exercises is to learn the material better.

Project grading

Each Project Exercise will be graded out of three, roughly as follows:

Project topic

15%
30%
55%

3/3 = essentially correct and complete, possibly with a few trivial errors
2/3 = minor errors or minor omissions or presentation problems
1/3 = major errors or major omissions or terrible presentation
0/3 = essentially wrong or totally incomplete or incomprehensible

The project this term deals with various balancing systems, such as that shown
below:

frictionless
pivots

plant

1
2

Rod #2

Rod #1

1
F

0
0
base
x
=
0

frictionless
0
interface
1
y
=
0

0
0
0

0
0.8889 1.333
x+
u
0
1
0

0
1.333 8.000
0 0 0
0 0
x+

u
0 1 0
0 0

1
0
0 3.2667
0
0
0 19.600

2 of 3

ECE 488 Tentative Lecture Schedule


1. Introduction [2 lectures]

Whats control engineering all about?


Whats new in this course?
Applications of control
A review of the control engineering design cycle
Some useful Matlab commands

2. Design for SISO SystemsPart 1: Classical techniques [5 lectures]

Signals and Systems Basics


Stability of Interconnected Systems and the Routh-Hurwitz Stability Test
Steady-state Calculations
Transient Response Calculations
Root Locus Methods
Nyquist Plots, Phase and Gain Margins, Lead and Lag Compensation
Introduction to the Small Gain Theorem
Introduction to Loopshaping

3. Design for SISO SystemsPart 2: Performance Limitations in Control [11 lectures]

Limitations on Stabilization (Youla parameterization)


The Sensitivity and Complementary Sensitivity Functions: Performance Limitations due to
Control Topology
Performance Limitations due to ORHP Poles (time domain and frequency domain)
Performance Limitations due to ORHP Zeros (time domain and frequency domain)
Performance Limitations if there are both ORHP Zeros and ORHP Poles

4. Design for MIMO SystemsPart 1: Use of transfer function techniques [7 lectures]

Basic MIMO concepts


Breaking a MIMO problem up into a series of SISO problems
Extending SISO Transfer Function Methods: MIMO Nyquist, etc.
Extending SISO Transfer Function Methods: MIMO Loopshaping
MIMO performance limitations

5. Design for MIMO SystemsPart 2: Use of state-space techniques [8 lectures]

Basic state-space facts and manipulations


Controllability, observability, stabilizability, and detectability
State-feedback control (pole placement)
Observers, state-feedback using observers, and the separation principle
Addition of integral control

6. Design for MIMO SystemsPart 3: Introduction to Optimal Control [3 lectures]

Motivation
LQR Optimal Control Problem
LQR Solution
Optimal Observers
Important Example

3 of 3

You might also like