On the frequency scaling in continuous-time modeling

R. Pintelon, I. Kollár

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

When identifying continuous-time systems in the Laplace domain, it is indispensable to scale the frequency axis to guarantee the numerical stability of the normal equations. Without scaling, identification in the Laplace domain is often impossible even for modest model orders of the transfer function. Although the optimal scaling depends on the system, the model, and the excitation signal, the arithmetic mean of the maximum and minimum angular frequencies in the frequency band of interest is commonly used as a good compromise. In this paper we show (i) that the optimal frequency scaling also strongly depends on the estimation algorithm, and (ii) that the median of the angular frequencies is a better compromise for improving the numerical stability than the arithmetic mean.

Original languageEnglish
Title of host publicationConference Record - IEEE Instrumentation and Measurement Technology Conference
EditorsS. Demidenko, R. Ottoboni, D. Petri, V. Piuri, D.C.T. Weng
Pages1586-1589
Number of pages4
Volume3
Publication statusPublished - 2004
EventProceedings of the 21st IEEE Instrumentation and Measurement Technology Conference, IMTC/04 - Como, Italy
Duration: May 18 2004May 20 2004

Other

OtherProceedings of the 21st IEEE Instrumentation and Measurement Technology Conference, IMTC/04
CountryItaly
CityComo
Period5/18/045/20/04

Fingerprint

Convergence of numerical methods
scaling
Continuous time systems
numerical stability
Frequency bands
Transfer functions
transfer functions
excitation

Keywords

  • Continuous-time
  • Frequency domain identification
  • Frequency scaling
  • Numerical conditioning

ASJC Scopus subject areas

  • Instrumentation

Cite this

Pintelon, R., & Kollár, I. (2004). On the frequency scaling in continuous-time modeling. In S. Demidenko, R. Ottoboni, D. Petri, V. Piuri, & D. C. T. Weng (Eds.), Conference Record - IEEE Instrumentation and Measurement Technology Conference (Vol. 3, pp. 1586-1589)

On the frequency scaling in continuous-time modeling. / Pintelon, R.; Kollár, I.

Conference Record - IEEE Instrumentation and Measurement Technology Conference. ed. / S. Demidenko; R. Ottoboni; D. Petri; V. Piuri; D.C.T. Weng. Vol. 3 2004. p. 1586-1589.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Pintelon, R & Kollár, I 2004, On the frequency scaling in continuous-time modeling. in S Demidenko, R Ottoboni, D Petri, V Piuri & DCT Weng (eds), Conference Record - IEEE Instrumentation and Measurement Technology Conference. vol. 3, pp. 1586-1589, Proceedings of the 21st IEEE Instrumentation and Measurement Technology Conference, IMTC/04, Como, Italy, 5/18/04.
Pintelon R, Kollár I. On the frequency scaling in continuous-time modeling. In Demidenko S, Ottoboni R, Petri D, Piuri V, Weng DCT, editors, Conference Record - IEEE Instrumentation and Measurement Technology Conference. Vol. 3. 2004. p. 1586-1589
Pintelon, R. ; Kollár, I. / On the frequency scaling in continuous-time modeling. Conference Record - IEEE Instrumentation and Measurement Technology Conference. editor / S. Demidenko ; R. Ottoboni ; D. Petri ; V. Piuri ; D.C.T. Weng. Vol. 3 2004. pp. 1586-1589
@inproceedings{7ffe9d995ea14c9fa9abe5da2f8cb348,
title = "On the frequency scaling in continuous-time modeling",
abstract = "When identifying continuous-time systems in the Laplace domain, it is indispensable to scale the frequency axis to guarantee the numerical stability of the normal equations. Without scaling, identification in the Laplace domain is often impossible even for modest model orders of the transfer function. Although the optimal scaling depends on the system, the model, and the excitation signal, the arithmetic mean of the maximum and minimum angular frequencies in the frequency band of interest is commonly used as a good compromise. In this paper we show (i) that the optimal frequency scaling also strongly depends on the estimation algorithm, and (ii) that the median of the angular frequencies is a better compromise for improving the numerical stability than the arithmetic mean.",
keywords = "Continuous-time, Frequency domain identification, Frequency scaling, Numerical conditioning",
author = "R. Pintelon and I. Koll{\'a}r",
year = "2004",
language = "English",
volume = "3",
pages = "1586--1589",
editor = "S. Demidenko and R. Ottoboni and D. Petri and V. Piuri and D.C.T. Weng",
booktitle = "Conference Record - IEEE Instrumentation and Measurement Technology Conference",

}

TY - GEN

T1 - On the frequency scaling in continuous-time modeling

AU - Pintelon, R.

AU - Kollár, I.

PY - 2004

Y1 - 2004

N2 - When identifying continuous-time systems in the Laplace domain, it is indispensable to scale the frequency axis to guarantee the numerical stability of the normal equations. Without scaling, identification in the Laplace domain is often impossible even for modest model orders of the transfer function. Although the optimal scaling depends on the system, the model, and the excitation signal, the arithmetic mean of the maximum and minimum angular frequencies in the frequency band of interest is commonly used as a good compromise. In this paper we show (i) that the optimal frequency scaling also strongly depends on the estimation algorithm, and (ii) that the median of the angular frequencies is a better compromise for improving the numerical stability than the arithmetic mean.

AB - When identifying continuous-time systems in the Laplace domain, it is indispensable to scale the frequency axis to guarantee the numerical stability of the normal equations. Without scaling, identification in the Laplace domain is often impossible even for modest model orders of the transfer function. Although the optimal scaling depends on the system, the model, and the excitation signal, the arithmetic mean of the maximum and minimum angular frequencies in the frequency band of interest is commonly used as a good compromise. In this paper we show (i) that the optimal frequency scaling also strongly depends on the estimation algorithm, and (ii) that the median of the angular frequencies is a better compromise for improving the numerical stability than the arithmetic mean.

KW - Continuous-time

KW - Frequency domain identification

KW - Frequency scaling

KW - Numerical conditioning

UR - http://www.scopus.com/inward/record.url?scp=4644265668&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4644265668&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:4644265668

VL - 3

SP - 1586

EP - 1589

BT - Conference Record - IEEE Instrumentation and Measurement Technology Conference

A2 - Demidenko, S.

A2 - Ottoboni, R.

A2 - Petri, D.

A2 - Piuri, V.

A2 - Weng, D.C.T.

ER -