| Week nr. |
Date |
Subject Tehme Topic |
| 34 |
22.8. |
Introduction, notations,
state space models, system theory, state space
analysis. Syllabus: Lecture notes ch 1.1, 1.5,
example 1.1.
Introduction, notations, dynamic models, system
theory, state space and transfer function model
analysis.
Lecture contents: Lecture
1
Video Lecture: Se link
(Unfortunately without sound !)
|
| 35 |
29.8
|
System theory and state
space analysis. Poles and zeroes. Linearizing
non-linear models. Lecture notes ch 1. Introduction
to the PID controller, Ch. 4
Lecture contents: Lecture
2
Video Lectures:
1) Fall 2016: Ch.1.12 linearization+Exercise 3
Lecture2
Erratum: Notice that the A, B and C matrices
should be as in Eq. (30) in oving_pid_reaktor.pdf
2) Fall 2017:
Lecture2a Lecture2b
|
| 36 |
5.9 |
System teory and state
space analysis. PI-controller.The Skogestad
method.. Lecture notes ch. 3, paper by
Skogestad. Exercise 3, linearizing the chemical
reactor model.
Lecture contents: Lecture
3
Video Lecture:
1) Fall 2016: Lecture
3
2) Fall 2017: Lecture3a
Lecture3b
MATLAB Example files:
Numerical linearization:
1) Calculate A: jacobi.m
fx_chemreac.m
2) Calculate both A and B: jacobi2.m fx_chemreac2.m
>> % Linearized model in Exercise 3 may be
calculated as
>> [A,B]=jacobi2('fx_chemreac2',0,[2.5;1],25)
3) Simulate step response of chemical reactor
model: main_reacsim.m
Notice: Calculate right hand side in
dot(x)=f(x,u): fx_chemreac.m
|
| 37 |
12.9 |
PID-controller:
Skogestads method ch. 3.1 lecture notes. The half
rule for model reduction. Direct design ch. 7.
Standard controllers (P, PI og PID) Ch. 4.
Discretization and implementation of PI og PID
controllers.
Lecture contents: Lecture
4
Video Lecture:
Fall 2016: Lecture
4 Lecture 4b
Fall 2017: Lecture4a
Lecture4b
m-file for PI control of the chemical reactor:
main_reacsim_pid.m
Implementing a PI controller: ex1_pi.m
Inverse response example: main_invresp_ex.m
|
| 38-39 |
19.9-26.9 |
Ch. 4.3 Anti windup and
constraints. Ch. 4.4 Bumpless transfer between
manual and automatic control. Direct design of PI
controllers, examples from ch.. 3, 4 og 5 in lecture
notes.
m-file examples from Ch 4: ex1_pi_anti.m
ex1_pi_bump.m
Video Lectures:
Fall 2016:
1) anti windup+bumpless transfer Lecture5
2) tuning PI controller ex.: Lecture5b
+ Lecture5c
Fall 2017: Lecture5a
Lecture5b
Example 3.3:
In laplace plane, using approx to time delay: ex33_laplace2.m
In time domain, exact implementation of time
delay: ex3b_half.m
If time: Controllability. Observability. Linear
transformation av models, Ch. 1, lecture notes.
|
| 39-40 |
26.9-3.10 |
P and PI controllers for
integrating plus time delay process, disturbance
rejection and in particular disturbance at the input
side.
Lecture Contents:
1) Lect notes Ch.
3.8+3.8.1.
2) Di
Ruscio (2010) paper Ch. 4.1+4.2
Video Lectures:
1) Fall16: Lecture6
Lecture6b
2) Fall17: Lecture6a
Lecture6b
Lecture6c
New PI controller tuning method for
integrating+time delay systems
|
40
|
3.10
|
1) Ch. 3.9.Retuning a PI controller
for oscillating feedback loop, integrating+time
delay process.
2) Exercise example:
exercise_retune_pi.pdf ("Increase KP to avoid
oscillations")
3) Integrating+time constant+delay plant, cascade
control and PD controller.
4) Final Exam 2016. Task1
m-file
Video Letures:
Fall 2016 Lecture7
Fall 2017: Lecture7_fall17
4) Exercise 4: Canonical forms and controller
design. Lecture notes ch 2. PD control and state
feedback, examples. |
| 41 |
9.10. |
No ordinary
Lecture
Suggestions:
1) Read Ch. 3-4 and the examples.
2) Work with Task 1 final Exam 2014.
|
|
42
|
17.10
|
Introduction to frequency alalysis. Ch. 9
Work with example: main_ex_step1.m
Video Lecture:
Fall 2016: Lecture8
(Theory from Ch. 9.2.)
Lecture8b (Work trough example.)
Lecture8c (Interpretation of GM and PM)
Fall2017: Lecture8a
Lecture8b
New PI controller tuning method
for integrating+time delay systems
|
| 43 |
24.10 |
PID-control, feedback
systems, feed forward control. Frequency
analysis.
Lecture notes, Ch. 6 (feedback systems)
, Ch. 8 (feed forward, ratio control) ,
Cascade control+ double int. ex., Ch. 9 (Frequency
analysis)
Sensitivity
index (Ms), Gain Margin (GM) and Phase
Margin (PM)
Video Lecture: Lecture9
+ Lecture9b
|
| 44 |
31.10 |
Frequency analysis. Phase
and gain margin. Bode's stability criterion. PID
control with Ziegler Nichols method. Lecture notes.
Video Lecture: Lecture10
+ Lecture10b
+ Lecture10c
Remarks:
Lecture10 (The Bode stability criterion)
Lecture10b (the Ziegler Nichols method)
Lecture10c (Ex. 2: SIMC tuning for 1sr order+time
delay plant)
MATLAB m-file for Ex.2 in Lecture10c main_sim_1st_margins.m
|
| 45-46 |
7.11-14.11 |
1) Non-Minimum phase
systems Ch. 9.6.
2) Bandwidt of a control system Ch. 9.7.
3) Margins using SIMC method, integrating pluss
time delay process with PI controller. Task.
Solution:
task_1511.m
Video Lecture : (above items 1)+2)+3) )
Lecture 11
4) Velocity/deviation form of the PI controller.
From Laplace formulation to continuous time , to
discrete time and to the velocity formulation.
Partly repetition from Ch. 4 and leading to the
result in Ch. 10.4 (10.4.1+10.4.2+10.4.3).
Video Lecture: Lecture11b
(Velocity PID formulation)
5) Example of how to implement the PI velocity
formulation: ex3b_half_pi_du.m
for controlling a process model: y=h_p(s)(u+v)
where h_p(s)=k*exp(-tau*s)/s with slope, k=1, and
time delay, tau=1
6) New method for tuning PI controllers for
integrating plus time delay systems: DiRuscio2010
Feedforward control. Lecture notes.
State estimation, modal estimation and control. Notes.
Exam task on black board. m-files to the notes: lpe.m, lpe2.m
|
| 46-47 |
14.11-21.11 |
The Smith Predictor
Video Lecture: Lecture12
Exercise: Example
task. Solution proposal:
main:solution_extask1.m Theory: Sensitivity
index
Pade approximation to time delay. Stability
analyzis. Feed forward control. Lecture notes, Ch
5 and 8.
D and PD controller for double integrator + time
delay process. Example hp_double_int.m
|
| 47 |
21.11 |
Discrete control.
Discrete PID-controller. Incremental form. Trapes
method. Stability. s-plane and z-plane. Ch.
10. Ratio control.
Exercise 10.
New method for tuning PI controllers for
integrating plus time delay systems: DiRuscio2010
|
| 47 |
21.11 |
Smith-predictor,
inverse response control, modified
Smith-predictor. System teory for discrete
systems if time.
Handwritten block-diagram reduction of the Smith
predictor. (lecture
notes)
New PI controller tuning method for
integrating+time delay systems
|
| 47 |
21.11 |
Multivariable control:
RGA-analysis and decoupling.
Video Lecture: Lecture13
(Pairing of variables in MIMO systems)
Summing up the syllabus:. PID-control, Skogestads
method, frequency analysis, Ziegler-Nichols
method, Bodes stability criterion, stability
analysis.
|
