Coursecode: wb2301
Coursename: System Identification A
DUT creditpoints: 5
ECTS creditpoints: 8 |
Subfaculty
of Mechanical Engineering and Marine Technology |
Lecturer(s): Lunteren, dr.ir. A. van |
Tel.: 015-2782126 |
Catalog
data:
Non-parametric system identification based on spectral density estimates. application in
open and closed loop systems. |
Courseyear:
3
Semester: 0/0/0/0/4
Hours p/w: 4
Other hours:
Assessment: See remarks
Assessm.period(s): See remarks
(see academic calendar) |
Prerequisites:
wb2307 |
Follow
up: wb2403/5 |
Detailed
description of topics:
Problems
in obtaining relations between spectral densities and transfer functions starting from
time domain derivations
A modified
Fourier Transform for stochastic processes.
Application
to relations between transfer functions and spectral densities in a closed loop and to
identification of multivariable systems in open and closed loops.
Fourier
Transform of sampled data by using a dirac comb. Shannon's Theorem. The Discrete Fourier
Transform (DFT) and the Fast Fourier Transform (FFT).
The
z-transform and its relation to the DFT.
Description
of discrete noise by means of its shaping filter. AR-, MA-, and ARMA-processes.
Estimation of
mean-lagged-product functions, covariance functions and correlation functions. Properties
of the estimators for these functions.
Estimation of
spectral densities according to the classical method based on transformation of covariance
function estimates; transformation windows.
Estimation of
spectral densities based on transformation of signals. Comparison with the classical
method.
Properties of
estimators for spectral densities, transfer functions and coherency functions.
|
Course
material:
Lecturenotes "Signaalanalyse'' (in Dutch), Van Lunteren / Dankelman. |
References
from literature:
Jenkins G.M. and Watts D.G. (1968). Spectral analysis
and its applications. Holden Day, San Francisco.
Priestly M.B. (1981). Spectral analysis and time
series. Academic Press, London. ISBN 0-12-564922-3.
|
Remarks
(specific information about assesment, entry requirements, etc.):
Assessment: combination of written and oral made by appointment one day every month |
Goals:
Being able to apply non-parametric system idetification techniques of which the
theoretical basis is given in course wb2307. Being able to recognise and avoid the
pitfalls inherent to working with estimators and taking into account that in general
measurements have to be executed in closed loop systems. |
Computer
use:
Excercises on a PC with a number of available
MATLAB-programs. |
Laboratory
project(s): |
Design
content: |
Percentage
of design: 0% |