Comparison of process identification techniques in the time and frequency domains

Tony Kealy; Aidan O'Dwyer

doi:10.21427/abxv-vc53

Comparison of process identification techniques in the time and frequency domains

Tony Kealy, Aidan O'Dwyer

Technological University Dublin

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This paper describes ten methods to identify a mathematical model for a real process with a time delay. The process is the Process Trainer, PT326 from Feedback Instruments Limited. Six of the methods use step response data and one of the methods uses impulse response data for identification. Two of the methods use frequency response data and the final method uses information from relay-based experiments. The best results are obtained using a combined analytical and gradient method [6] in the frequency-domain and, in the time-domain, using the two-point algorithm [1] and a method proposed by Suganda et al. [5].

Original language	English
Title of host publication	Proceedings of ICSE 2003 (Sixteenth International Conference on Systems Engineering)
Publisher	Coventry University
Pages	350-354
DOIs	https://doi.org/10.21427/abxv-vc53
Publication status	Published - 2003
Event	Sixteenth International Conference on Systems Engineering - Coventry, United Kingdom Duration: 9 Sep 2003 → 11 Sep 2003

Conference

Conference	Sixteenth International Conference on Systems Engineering
Country/Territory	United Kingdom
City	Coventry
Period	9/09/03 → 11/09/03

Keywords

process identification
time delay
step response data
impulse response data
frequency response data
relay-based experiments
analytical method
gradient method
two-point algorithm

Access to Document

10.21427/abxv-vc53Licence: CC BY-NC-SA

Cite this

@inproceedings{1de1926c0f8e4cd591ca7e82db845169,

title = "Comparison of process identification techniques in the time and frequency domains",

abstract = "This paper describes ten methods to identify a mathematical model for a real process with a time delay. The process is the Process Trainer, PT326 from Feedback Instruments Limited. Six of the methods use step response data and one of the methods uses impulse response data for identification. Two of the methods use frequency response data and the final method uses information from relay-based experiments. The best results are obtained using a combined analytical and gradient method [6] in the frequency-domain and, in the time-domain, using the two-point algorithm [1] and a method proposed by Suganda et al. [5].",

keywords = "process identification, time delay, step response data, impulse response data, frequency response data, relay-based experiments, analytical method, gradient method, two-point algorithm",

author = "Tony Kealy and Aidan O'Dwyer",

year = "2003",

doi = "10.21427/abxv-vc53",

language = "English",

pages = "350--354",

booktitle = "Proceedings of ICSE 2003 (Sixteenth International Conference on Systems Engineering)",

publisher = "Coventry University",

note = "Sixteenth International Conference on Systems Engineering ; Conference date: 09-09-2003 Through 11-09-2003",

}

TY - GEN

T1 - Comparison of process identification techniques in the time and frequency domains

AU - Kealy, Tony

AU - O'Dwyer, Aidan

PY - 2003

Y1 - 2003

N2 - This paper describes ten methods to identify a mathematical model for a real process with a time delay. The process is the Process Trainer, PT326 from Feedback Instruments Limited. Six of the methods use step response data and one of the methods uses impulse response data for identification. Two of the methods use frequency response data and the final method uses information from relay-based experiments. The best results are obtained using a combined analytical and gradient method [6] in the frequency-domain and, in the time-domain, using the two-point algorithm [1] and a method proposed by Suganda et al. [5].

AB - This paper describes ten methods to identify a mathematical model for a real process with a time delay. The process is the Process Trainer, PT326 from Feedback Instruments Limited. Six of the methods use step response data and one of the methods uses impulse response data for identification. Two of the methods use frequency response data and the final method uses information from relay-based experiments. The best results are obtained using a combined analytical and gradient method [6] in the frequency-domain and, in the time-domain, using the two-point algorithm [1] and a method proposed by Suganda et al. [5].

KW - process identification

KW - time delay

KW - step response data

KW - impulse response data

KW - frequency response data

KW - relay-based experiments

KW - analytical method

KW - gradient method

KW - two-point algorithm

U2 - 10.21427/abxv-vc53

DO - 10.21427/abxv-vc53

M3 - Conference contribution

SP - 350

EP - 354

BT - Proceedings of ICSE 2003 (Sixteenth International Conference on Systems Engineering)

PB - Coventry University

T2 - Sixteenth International Conference on Systems Engineering

Y2 - 9 September 2003 through 11 September 2003

ER -

Comparison of process identification techniques in the time and frequency domains

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this