Best conditioned parametric identification of transfer function models in the frequency domain

Author

Rolain, Y. ; Pintelon, R. ; Xu, K.Q. ; Vold, H.

Author_Institution

Vrije Univ., Brussels, Belgium

Volume

40

Issue

11

fYear

1995

fDate

11/1/1995 12:00:00 AM

Firstpage

1954

Lastpage

1960

Abstract

It is shown that rational transfer function models based on orthogonal Forsythe polynomials minimize the condition number of the Jacobian of estimators in a least-squares framework. As a result, very high order linear time-invariant systems can be identified. The numerical stability of the estimation of the parameters and their derived quantities (zeros, poles, …) are obtained. Statistical uncertainty bounds are provided. The method is illustrated on a 100th order simulated system and a 120th order measured beam-structure

Keywords

least squares approximations; numerical stability; parameter estimation; polynomials; transfer functions; best conditioned parametric identification; condition number; frequency domain; least-squares estimators; numerical stability; orthogonal Forsythe polynomials; rational transfer function models; statistical uncertainty bounds; very high order linear time-invariant systems; Cost function; Frequency domain analysis; Frequency estimation; Frequency measurement; Frequency response; Parameter estimation; Poles and zeros; Polynomials; Transfer functions; Virtual manufacturing;

fLanguage

English

Journal_Title

Automatic Control, IEEE Transactions on

Publisher

ieee

ISSN

0018-9286

Type

jour

DOI

10.1109/9.471223

Filename

471223