Unbiased identification of continuous-time parametric models using a time-weighted integral transform

Shyh-Hong Hwang, Min Lang Lin

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

A time-weighted integral transform is presented to identify a continuous SISO or MIMO parametric model based on a single dynamic test under open-loop or closed-loop operation. Moving-horizon algorithms are proposed to obtain unbiased estimates of the model parameters. The off-line algorithm in a least-squares form and the on-line algorithm in a recursive form are provided. An effective technique based on pattern recognition is also developed to determine the system order and time delay from observed data in a simple manner. Furthermore, the proposed method can be easily applied as a model reduction technique that results in an ideal model with delay for any specified order.

Original languageEnglish
Pages (from-to)1170-1199
Number of pages30
JournalChemical Engineering Communications
Volume190
Issue number9
DOIs
Publication statusPublished - 2003 Sep 1

Fingerprint

Identification (control systems)
MIMO systems
Pattern recognition
Time delay

All Science Journal Classification (ASJC) codes

  • Chemistry(all)
  • Chemical Engineering(all)

Cite this

@article{554353157d354be7955b74ff20215393,
title = "Unbiased identification of continuous-time parametric models using a time-weighted integral transform",
abstract = "A time-weighted integral transform is presented to identify a continuous SISO or MIMO parametric model based on a single dynamic test under open-loop or closed-loop operation. Moving-horizon algorithms are proposed to obtain unbiased estimates of the model parameters. The off-line algorithm in a least-squares form and the on-line algorithm in a recursive form are provided. An effective technique based on pattern recognition is also developed to determine the system order and time delay from observed data in a simple manner. Furthermore, the proposed method can be easily applied as a model reduction technique that results in an ideal model with delay for any specified order.",
author = "Shyh-Hong Hwang and Lin, {Min Lang}",
year = "2003",
month = "9",
day = "1",
doi = "10.1080/00986440302162",
language = "English",
volume = "190",
pages = "1170--1199",
journal = "Chemical Engineering Communications",
issn = "0098-6445",
publisher = "Taylor and Francis Ltd.",
number = "9",

}

Unbiased identification of continuous-time parametric models using a time-weighted integral transform. / Hwang, Shyh-Hong; Lin, Min Lang.

In: Chemical Engineering Communications, Vol. 190, No. 9, 01.09.2003, p. 1170-1199.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Unbiased identification of continuous-time parametric models using a time-weighted integral transform

AU - Hwang, Shyh-Hong

AU - Lin, Min Lang

PY - 2003/9/1

Y1 - 2003/9/1

N2 - A time-weighted integral transform is presented to identify a continuous SISO or MIMO parametric model based on a single dynamic test under open-loop or closed-loop operation. Moving-horizon algorithms are proposed to obtain unbiased estimates of the model parameters. The off-line algorithm in a least-squares form and the on-line algorithm in a recursive form are provided. An effective technique based on pattern recognition is also developed to determine the system order and time delay from observed data in a simple manner. Furthermore, the proposed method can be easily applied as a model reduction technique that results in an ideal model with delay for any specified order.

AB - A time-weighted integral transform is presented to identify a continuous SISO or MIMO parametric model based on a single dynamic test under open-loop or closed-loop operation. Moving-horizon algorithms are proposed to obtain unbiased estimates of the model parameters. The off-line algorithm in a least-squares form and the on-line algorithm in a recursive form are provided. An effective technique based on pattern recognition is also developed to determine the system order and time delay from observed data in a simple manner. Furthermore, the proposed method can be easily applied as a model reduction technique that results in an ideal model with delay for any specified order.

UR - http://www.scopus.com/inward/record.url?scp=0242317881&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0242317881&partnerID=8YFLogxK

U2 - 10.1080/00986440302162

DO - 10.1080/00986440302162

M3 - Article

VL - 190

SP - 1170

EP - 1199

JO - Chemical Engineering Communications

JF - Chemical Engineering Communications

SN - 0098-6445

IS - 9

ER -