How Students Are Graded

Evaluation and Grading of Advanced Control
Systems Course
1 Grading
The final grading of the Advanced Control Systems course is based on written exam (40 points), TP
reports (30 points) and evaluation meeting (30 points). The final grade then is computed using the
following table.
Points 96-100 91-95 86-90 61-65 56-60 51-55 6-10 1-5 0
Grade 6.00 5.75 5.50 4.25 4.00 3.75 1.50 1.25 1.00
2 Written Exam
The exam is open book and open notes. The use of scientific calculator is allowed but not the computer.
The questions and problems are limited to Chapter 1 and Chapter 3 of the course-notes. The learning
outcomes of each chapter are evaluated by the problems and questions that are detailed as follows.
2.1 Chapter 1
The student should be able to
Compute any norm of a given signal or system (see Example 1.2 and Example 1.3).
Example: Compute the 1-norm, 2-norm and -norm of

2 0 t 10
u(t) =
0 elsewhere
Analyze the internal stability of a closed-loop system.

Convert a specific closed-loop performance to a norm constraint for a weighted closed-loop sensitivity
function or to minimization of an objective function.
Example: Consider the following closed-loop system:
v(t)
r(t) e(t) u(t)
- ?
h y(t)
+
- h - K - G
-
6
Which criterion should be minimized to minimize the two norm of the output when
r(t) = 0 and (1) v(t) is a Dirac impulse signal, (2) v(t) is a step signal, (3) v(t) = sin 0 t,
(4) v(t) is a bounded two-norm signal whose energy is concentrated between 1 and 2 .
1
Convert parametric and multimodel uncertainty to unstructured frequency-domain uncertainty (i.e.
additive, multiplicative, feedback uncertainty). See Examples 1.6, 1.7, 1.8 and 1.9.
Find the limit of achievable performance for unstable and non-minimum phase systems.
Example: Given
s1 k
G(s) = and W1 (s) =
s2 s+1
find for which values of k > 0 the nominal performance W1 S < 1 can be achieved.
Solution: We have z = 1 and p = 2 and according to the maximum modulus theorem:

z + p k 3
W1 S > W1 (z) = 0 < k < 2/3
z p 2 1
2.2 Chapter 3
The student should be able to design RST digital controllers
with specifications on desired closed-loop poles.
with fixed terms in the controller,
for internal model control, model reference control and minimum variance control,
using Q parameterization for robustness.
See Example 3.5, 3.7 and the following examples:
Example 1: Consider the following plant model:
0.5q 1 0.4q 2
G(q 1 ) =
1 1.2q 1 + 0.35q 2
Design an IMC controller with integrator and the same dynamics for tracking and regu-
lation.
0.5(z 0.8)
G(z) =
(z 0.7)(z 0.5)
Take Pd (q 1 ) = 1 0.6q 1 and the same dynamics for tracking and regulation. Compute
R0 (q 1 ), S0 (q 1 ) and T0 (q 1 ). Add HR (q 1 ) = 1 + q 1 and an integrator into the
controller using Q parameterization.
0.5q 1 0.4q 2
G(q 1 ) =
1 1.2q 1 + 0.35q 2
Take Pd (q 1 ) = 10.6q 1 . Design a Model Reference Controller. Add HR (q 1 ) = 1+q 1

and an integrator into the controller using Q parameterization.
Example 4: Consider the following ARMAX plant model:
y(t + 1) = (1.2q 1 0.35q 2)y(t) + (0.5 0.4q 1 )u(t) 0.6e(t) + e(t + 1)
Design a Minimum Variance Controller with integrator action.
2
3 TP Reports
Every group of two students prepares one report per TP and submits it on the moodle in due time. Each
TP report will be evaluated before the evaluation meeting based on the completeness and correctness of
the results and the clarity of presentation. Each report has maximum 15 points which is given equally to
both students. The grade, however, is provisional and will be reduced after the evaluation meeting if it
turns out that the students (or one of them) are not aware of what is written in the report.
4 Evaluation Meeting
The evaluation meeting will be planned individually during the exam period via moodle. Each meeting
will last about 15 minutes. During the meeting the students can only use their reports (no access to
the course-notes). The objective of this meeting is to evaluate the level of understanding of the students
about the subject of the TPs and the course (especially Chapter 2 and Chapter 4).
4.1 TP questions:
Assume that a student has used the following command in his/her report:
performance=conper(LS,[Mm,Ku,wh],wc/s);
The student may be asked the following questions:
Explain the loop shaping method, which criterion is minimized?
Why the desired loop is chosen as wc/s?
What is the definition of the modulus margin Mm?
What are Ku and wh? How they are chosen? How they are implemented as constraints in the
optimization problem?
4.2 Theoretical questions:

Since Chapter 4 is not evaluated in the written exam and TPs, some general questions are asked during
the evaluation meeting. The questions are very general and limited to the following list:
1. Parametric Adaptation Algorithm:
Explain the general structure of a PAA.
What is a PAA with filtering, with dead-zone, with projection, with normalization?
For a PAA with the following recursive adaptation gain
F 1 (t + 1) = 1 (t)F 1 (t) + 2 (t)(t)T (t)
Explain the decreasing gain, constant forgetting factor, variable forgetting factor and constant
trace.
2. Adaptive Control:
What is an adaptive control system. Give its general scheme and compare it with conventional
and robust control.
Compare direct and indirect adaptive control.
Give the block-diagram of Model Reference Adaptive Control.
3
Explain the certainty equivalence and separation principles.
Explain three strategies for the implementation of the indirect adaptive control.
3. Switching Adaptive Control:
Give the block-diagram of the switching adaptive control and compare it with gain-scheduled
control.
Explain the monitoring signal and the switching logic.
Discuss the stability of switching adaptive control for the trivial case and nontrivial case.
What is the injected system and in which condition the injected system is stable?

How Students Are Graded

Uploaded by

Copyright:

Available Formats

How Students Are Graded

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

How Students Are Graded

Uploaded by

Copyright:

Available Formats

Evaluation and Grading of Advanced Control

Example: Compute the 1-norm, 2-norm and -norm of

Analyze the internal stability of a closed-loop system.

Example: Consider the following closed-loop system:

Solution: We have z = 1 and p = 2 and according to the maximum modulus theorem:

Example 1: Consider the following plant model:

Take Pd (q 1 ) = 10.6q 1 . Design a Model Reference Controller. Add HR (q 1 ) = 1+q 1

y(t + 1) = (1.2q 1 0.35q 2)y(t) + (0.5 0.4q 1 )u(t) 0.6e(t) + e(t + 1)

Design a Minimum Variance Controller with integrator action.

4.2 Theoretical questions:

F 1 (t + 1) = 1 (t)F 1 (t) + 2 (t)(t)T (t)

You might also like