You need JavaScript to view this

Algorithms and tools for system identification using prior knowledge

Thesis/Dissertation:

Abstract

One of the hardest problems in system identification is that of model structure selection. In this thesis two different kinds of a priori process knowledge are used to address this fundamental problem. Concentrating on linear model structures, the first prior advantage of is knowledge about the systems` dominating time constants and resonance frequencies. The idea is to generalize FIR modelling by replacing the usual delay operator with discrete so-called Laguerre or Kautz filters. The generalization is such that stability, the linear regression structure and the approximation ability of the FIR model structure is retained, whereas the prior is used to reduce the number of parameters needed to arrive at a reasonable model. Tailorized and efficient system identification algorithms for these model structures are detailed in this work. The usefulness of the proposed methods is demonstrated through concrete simulation and application studies. The other approach is referred to as semi-physical modelling. The main idea is to use simple physical insight into the application, often in terms of a set of unstructured equations, in order to come up with suitable nonlinear transformation of the raw measurements, so as to allow for a good model structure. Semi-physical modelling is less ``ambitious`` than physical  More>>
Authors:
Publication Date:
Sep 01, 1994
Product Type:
Thesis/Dissertation
Report Number:
LIU-TEK-LIC-1994-42
Reference Number:
SCA: 990200; PA: AIX-26:006790; EDB-95:011666; SN: 95001299400
Resource Relation:
Other Information: TH: Thesis (TeknL).; PBD: Sep 1994
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; KNOWLEDGE BASE; SYSTEMS ANALYSIS; ALGORITHMS; BWR TYPE REACTORS; COMPUTERIZED SIMULATION; FUNCTIONS; NEUTRON FLUX; REGRESSION ANALYSIS; 990200; MATHEMATICS AND COMPUTERS
OSTI ID:
10102782
Research Organizations:
Linkoeping Univ. (Sweden). Dept. of Electrical Engineering
Country of Origin:
Sweden
Language:
English
Other Identifying Numbers:
Journal ID: ISSN 0280-7971; Other: ON: DE95610168; ISBN 91-7871-422-2; TRN: SE9400186006790
Availability:
OSTI; NTIS; INIS
Submitting Site:
SWDN
Size:
143 p.
Announcement Date:
Jun 30, 2005

Thesis/Dissertation:

Citation Formats

Lindskog, P. Algorithms and tools for system identification using prior knowledge. Sweden: N. p., 1994. Web.
Lindskog, P. Algorithms and tools for system identification using prior knowledge. Sweden.
Lindskog, P. 1994. "Algorithms and tools for system identification using prior knowledge." Sweden.
@misc{etde_10102782,
title = {Algorithms and tools for system identification using prior knowledge}
author = {Lindskog, P}
abstractNote = {One of the hardest problems in system identification is that of model structure selection. In this thesis two different kinds of a priori process knowledge are used to address this fundamental problem. Concentrating on linear model structures, the first prior advantage of is knowledge about the systems` dominating time constants and resonance frequencies. The idea is to generalize FIR modelling by replacing the usual delay operator with discrete so-called Laguerre or Kautz filters. The generalization is such that stability, the linear regression structure and the approximation ability of the FIR model structure is retained, whereas the prior is used to reduce the number of parameters needed to arrive at a reasonable model. Tailorized and efficient system identification algorithms for these model structures are detailed in this work. The usefulness of the proposed methods is demonstrated through concrete simulation and application studies. The other approach is referred to as semi-physical modelling. The main idea is to use simple physical insight into the application, often in terms of a set of unstructured equations, in order to come up with suitable nonlinear transformation of the raw measurements, so as to allow for a good model structure. Semi-physical modelling is less ``ambitious`` than physical modelling in that no complete physical structure is sought, just combinations of inputs and outputs that can be subjected to more or less standard model structures, such as linear regressions. The suggested modelling procedure shows a first step where symbolic computations are employed to determine a suitable model structure - a set of regressors. We show how constructive methods from commutative and differential algebra can be applied for this. Subsequently, different numerical schemes for finding a subset of ``good`` regressors and for estimating the corresponding linear-in-the-parameters model are discussed. 107 refs, figs, tabs.}
place = {Sweden}
year = {1994}
month = {Sep}
}