SC4070 - Control Systems Lab: Course Information For Academic Year 2012/2013, Q3

Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

SC4070 Control Systems Lab

Course Information for Academic Year 2012/2013, Q3

General Information

Course year

Msc programs

Term

third quarter (Feb 11th to the end of March) This course, provided by the Delft Center for Systems and
Control, TU Delft, is intended for MSc (graduate-level) stu4 (112 hours)
dents who already have basic knowledge of control thean undergraduate feedback control course, ory (at the BSc-level as taught, for instance, at Electrical,
Mechanical or Aerospace Engineering and Applied Maths).
experience with Matlab, Simulink
The main benefit from this course will be the hands-on ex4 lectures in February:
perience with the design and implementation of a computercontrolled system on a real laboratory-scale process. RelaMon 11th, Fri 15th, Fri 22nd, Mon 25th
tively abstract parts of control theory will become clear and
1 lab demo:
tangible.

ECTS credits
Prerequisites

Lectures

Mon February 18th


Lab sessions1

week 9, 10 and 11 in 3mE 8f-0-420,

Students will be working together with the same lab setups.


The lab projects are carried out in groups of three students
(more details are given further in this text).

Mekelweg 2
Lecturer

dr. ir. Alessandro Abate


Delft Center for Systems and Control
tel: +31-15-27 85606
email: [email protected]

Lab Assistants

ir. Hans Yoo


Delft Center for Systems and Control
tel: +31-15-27 84106
email: [email protected]
ir. Hildo Bijl
Delft Center for Systems and Control
tel: +31-15-27 81362
email: [email protected]

Examination

written report, final presentation


evaluating results of the lab sessions

WWW page

http://www.dcsc.tudelft.nl/ sc4070

Blackboard

search course SC4070 on


http://blackboard.tudelft.nl

There are three sessions of about 4 hours each. The


exact days and times will be scheduled during the first
lectures.

Course Subject
Practically all modern control systems are based on digital
computers and specialized micro-controllers. By using digital control, better performance and improved functionality
can be obtained than with analog systems.
The goal of this course is to gain hands-on experience with
the design and implementation of a computer-controlled
system. We will use a discrete-time approach, in which the
system to be controlled is modeled both by discretizing an
available continuous-time physical model and by using system identification. A systematic, MATLAB-supported design methodology is followed, using a state estimator (observer) and a state-feedback controller.
In the first two weeks, four lectures are given in order to refresh the theoretical and methodological background. Furthermore, a lab demo will introduce the students to the
available experimental setups. Then, the students work in
groups of three in the lab, with a setup of their choice (inverted pendulum, helicopter model, inverted wedge, or
the container crane model). The mathematical model of
the process to be controlled is provided. The assignment is
to be stated in terms of the control objective. The results
will be summarized in a report and a final presentation will
be given. The grade is determined on the basis of lab participation, the report and the presentation.
Keywords: Computer-controlled systems, analysis of
discrete-time systems, practical design of sampled-data
controllers, implementation of controllers in M ATLAB and
Simulink.

Instructional Objectives
After successfully completing the course, you should be
able to:
- perform system identification or parameter estimation
of the setup at hand,
- develop a model of a (nonlinear) dynamic system in
Simulink, linearize (trim) and discretize the model,
- select properly the sampling period, given specifications for the closed-loop performance,
- design a discrete-time controller and observer, using
effectively the Control Systems Toolbox of M ATLAB,
- implement the controller in Simulink and interface it
to the physical process with the help of the Real-Time
toolbox,
- evaluate the performance of the closed loop and modify the design parameters if necessary.

Course Material
The material includes electronic copies of the transparencies used in the lectures, M ATLAB demos and examples.
Books Computer Controlled Systems 3rd ed.
by
K.J. Astrom and B. Wittenmark (Prentice Hall, 1997)
can be consulted, mainly Chapters 1 through 9 are
of interest for this course. A number of supplements
(errata list, solutions manual, etc.) are available from
the course WWW page and also from the authors at
http://www.control.lth.se.
Feedback Control of Dynamic Systems 5th ed. by Franklin,
Powell, and Emami-Naeini. Prentice Hall, 2006.
Slides in PostScript and PDF formats are available for
download from the course WWW page. You may want to
print the transparencies before each lecture and use them
for your notes.
MATLAB scripts and demos can be downloaded from
the course WWW page. These will help you to gain
more insight both in the theoretical matter and the
implementation.

2. Implement a Simulink model of the setup as your


home preparation. Prepare also an experiment to
identify the parameters of the system and test the
experiment on the Simulink model (week 7 and 8).
3. Calibrate the model to approximately match the
process. Carry out identification experiments, collect
data (lab sessions, weeks 9-11).
4. Estimate model parameters through system identification (lab sessions, weeks 9-11).
5. Design a controller for the simulation model as your
home preparation.
6. Test and fine-tune the controller on the process (lab
sessions, weeks 9-11).
7. Write the final report (max 10 pages), prepare the
presentation.
8. Present the results (presentations will take place in
week 12 of Q3 week of March 18th).
9. Write report (week 12 and week 13). Deadline: hand
in a printed copy by Friday AM March 29th, 2013.
One lab session takes four hours, the exact days and times
will be scheduled during the first two lectures, taking into
account your preferences and constraints. If the schedule
permits, you may, of course, spend in the lab as much
time as you wish. Let us stress that it makes little sense
to come to the lab sessions unprepared. For this course it
holds perhaps more than for other ones that what you get
out of it is directly proportional to what you yourself put in.
Matlab and Simulink. One of the objectives is to learn to
use M ATLAB, Simulink and the Control Systems Toolbox
for computer-assisted design, analysis and implementation
of control systems. It is important that you can use this
software for your home preparation for the lab sessions.
Matlab and Simulink including the Control Systems
Toolbox is available via blackboard. If you have no home
PC available you can use the computers located in the
labs, discuss this with the lecturer. Make sure that the
Matlab or Simulink models that you create on your home
PC are compatible with the Matlab version installed at the
lab-machines (R2012B).

Course Method
The main ingredients of this course are lectures, homework preparation, practical laboratory sessions, reporting
and presentation.
Lectures. There are four lectures in the first two weeks of
the quarter, two lectures per week. The purpose of the lectures is to refresh the relevant theory and methodology.
Lab demo. Will be given in the first two weeks and will
introduce the experimental setups to the students.
Laboratory sessions are the most essential part of this
course. The purpose is to gain hands-on experience with
the control of a real physical process. The lab sessions have
the following structure:
1. Choose your two partners and the laboratory setup
(first week of education, week 7) The work is done
in groups of three: email names, student id and
email addresses to the Instructor .

Final grade. The grade is determined on the basis of lab


participation, of your report (one report per group), and of
the final presentation. Both the report and the presentation
are in English.
The main criteria for the report are completeness and clarity. On the basis of the report, any educated reader should
be able to reproduce your results with the laboratory setup.
All choices made (such as the selection of the sampling period, design parameters, etc.) must be properly motivated.
Evaluate critically the results obtained. Deliver the M ATLAB / Simulink files you developed per e-mail to the Lab
Assistants. The report is max 10 pages, and must be delivered on paper. Please indicate exact contributions of
each of the students in the group on the report.
The presentation should take 20 minutes in total, 15 minutes the presentation itself and 5 minutes for questions both

by fellow students and the lecturers. It is important that


each member of the group gives a part of the presentation
(approximately 7 minutes). A beamer will be available.
The suggested (approximate) distribution of the study load
over the term is given in the following table.
Week

Lectures

Home prep. 10

10

Laboratory

10

11

10

10

46

15

Report
14

14

13 Total
8

Presentation

Total

12

15

15

19

11

12

20

11

12

100

Questions? You are welcome to ask questions and discuss


problems before or after each lecture or during the laboratory sessions. Please, make an appointment with
the lecturer for theoretical questions or questions on
models;
the lab assistants for questions on models or on lab
experiments.

You might also like