|
|
|||||||||||||||||
Course Catalog 2013-2014
ASE-5056 Optimal and Robust Control with Matlab, 8 cr |
Additional information
Suitable for postgraduate studies
Person responsible
Terho Jussila, Risto Ritala, Robert Piche
Lessons
| Study type | P1 | P2 | P3 | P4 | Summer | Implementations | Lecture times and places |
|
|
|
|
|
|
|
|
|
Requirements
Written exam (or a set of written sub-exams) plus a PC Exam plus 6 PC sessions (each of 120 minutes) plus a practical LAB of 2-3 hours. In the last weeks of the course there are seminars to reduce the exams: An active participator of any seminar can skip exam questions on the seminar topic.
Completion parts must belong to the same implementation
Principles and baselines related to teaching and learning
Study briefly the topic/content of each session before the session, follow the session actively, present questions, try to solve problems before the exercise session, repeat the contents on own time after the session. Increase your knowledge and understanding gradually all the time! Use PC exercises for illustrations/repetitions of issues studied before them.
Learning Outcomes
After the course the student should be able to use Matlab, Simulink, Control System Toolbox and Symbolic Toolbox to design a H2/LQ (Linear Quadratic) and LQG (LQ Gaussian) control law for a LTI (Linear Time-Invariant) state space system, analyze robust stability of the overall system in both classical and modern ways; can propose implementations; can use Matlab and the toolboxes for both time domain modeling, simulation and various other analyses of various MIMO LTI systems, including nominal and robustness analyses; knows several quadratic performance indices and their properties; understands special features and challenges of MIMO control design; can use both state space and transfer function models in an appropriate way and make conversions between model types; can create simple Matlab tools when Toolboxes are not providing suitable ones.
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Time domain simulation of various model types with Simulink, Control System Toolbox and tools of the core Matlab. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | S functions of Simulink. Creating and masking sub-systems. Use of ode45 and dde23. |
| 2. | Matlab and Control System Toolbox for effective modelling: equilibriums, linearization, least squares, model reduction, model type conversions, building models from sub-system models. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | Solvers and optimizers. |
| 3. | Basics of Linear Time Invariant models of MIMO systems: example responses, stability, controllability, observability, transfer function matrix, frequency response. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | |
| 4. | Quadratic performance indices and signal norms, even for time-delay systems and including time-weighting. Linear Quadratic deterministic state-feedback control and quadratic optimal parametric control. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | |
| 5. | Classical and modern studies of robust stability: stability margins for MIMO systems and studies of unstructured uncertainty. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | |
| 6. | Vector random processess in time and frequency domain. Identification, spectral factorization. Mean and variance calculus, variance minimization. Stochastic regulator, Kalman filtering, LQG control. | Both DT (Discrete-Time) and CT (Continuos-Time) studies are available so that one of them is primary, the other is secondary. The decision will be made for every implementation separately. | |
| 7. | Improving reliability of the computations. |
Instructions for students on how to achieve the learning outcomes
Usual.
Assessment scale:
Numerical evaluation scale (1-5) will be used on the course
Partial passing:
Study material
| Type | Name | Author | ISBN | URL | Edition, availability, ... | Examination material | Language |
| Book | Nomen Nescio | The main course book will be specified separately for each particular implementation. | Yes | English | |||
| Summary of lectures | Lecturer | The lecture notes/slides will be specified separately for each particular implementation. | Yes | English |
Prerequisites
| Course | Mandatory/Advisable | Description |
| ASE-1130 Automation | Mandatory | 1 |
| ASE-1230 Basic Course in Systems Technology 2 | Mandatory | 1 |
| ASE-1250 Control of Dynamic Systems | Mandatory | 1 |
| ASE-1256 Introduction to Control and Automation | Mandatory | 1 |
| ASE-1257 Introduction to Control | Mandatory | 1 |
1 . One of the ASE courses 1230, 1250, 1256, 1257 is necessary if the student is not mastering the content of such a course otherwise.
Additional information about prerequisites
An extra supporting session on discrete-time models is available a day before the course start: see POP events of the course implementation.
Prerequisite relations (Requires logging in to POP)
Correspondence of content
| Course | Corresponds course | Description |
|
|
|
|
|
|
|
|
|
|
|
More precise information per implementation
| Implementation | Description | Methods of instruction | Implementation |
| Lectures Excercises Practical works Laboratory assignments Other contact teaching |
Contact teaching: 45 % Distance learning: 0 % Self-directed learning: 55 % |