Empirical Modelling 1ET100

Empirisk Modellering

Additional course material and old exams to download will later be available from here!

The course starts 10.15, Jan 18 2010 in room 1211. Welcome!

Teachers:

Bengt Carlsson (Examiner, lectures)
Johannes Nygren (Lab and project superv.)
Björn Halvarsson (Lab and project superv.)

Course literature:

Modellbygge och Simulering, Ljung och Glad, Studentlitteratur, 2004. The book is also available in English: "Modeling of Dynamic systems", Ljung and Glad, Prentice Hall.

Errata to course book

Kursplan, version 10-01-19

Goal:

To give knowledge on modelling of dynamic systems using measurement data (system identification). Give knowledge on how to model some energy processes.

Prerequisites:

Automatic control, Systems analysis.

Contents:

Empirical modelling of dynamic systems.
Linear regression, the least squares method and prediction error methods.
Simulation and optimisation.
Modelling strategies used in process industry; case-studies with guest lectures.
Project: Modelling and control of a some energy related processes.

Examination:

Lecture notes to download:

F1-Kap 1-2
F1-Kap 3

F2 Linjär regression

F3 Tidsdiskreta system

F4-F6 Beskrivning av signaler i frekvensdomänen

Extra: Stokastika processer - Sammanfattning av samband

F5 - Kap 10

F5 - Kap 11

F6-7 - Kap 12

F10 - Kap 14

F10 - Kap 13 mm

Recursive identification - Corrected 10-02-12

Problem solving sessions (L) to download:

Uppgifter räkneövningar

Compulsory Homeworkwork Assignment:

Computer laboratory files to download:

BL1
BL2
BL4
BL5 (Compulsory)

m-files, needed to run the excercises at home

Solutions BL1
Solutions BL2
Solutions BL4

Process laboratory files to download:

Process lab - Fan (Compulsory)
Wed 10 Feb, 09.00 : Bokningslistan sätts upp på dörren till sal 2314 =

Individual Task (Black box modeling):

Data for the task (blackbox.mat)
Data for the task in ASCII format (use only if you can not load the data above)

Group Project Task:

Alternative project to "Solhuset": Estimating the oxygen dynamics in an activated sludge proces, this alternative is for students with WWTP and Advanced Control background

Solhuset - solardat.mat
Vindddata - vind.mat
Extra task - eldata.mat

Project F.A.Q.

Answers to some frequently asked questions can be found here: F.A.Q.

MATLAB System Identification Toolbox: Dokumentation

Här hittar ni nyttig hjälp om SI-toolboxen och alla dess funktioner:
http://www.mathworks.com/access/helpdesk/help/toolbox/ident