Fault diagnosis based on black-box models with application to a liquid-level system

Luís Brito Palma; Fernando Vieira do Coito; Rui Neves Silva

doi:10.1109/ETFA.2003.1248772

Fault diagnosis based on black-box models with application to a liquid-level system

Luís Brito Palma, Fernando Vieira do Coito, Rui Neves Silva

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

6 Citations (Scopus)

Abstract

This paper proposes an on-line robust approach to fault detection and isolation (FDI) of dynamic systems. This FDI approach is based on black-box models: artificial neural networks (ANNs) and the autoregressive with exogenous input (ARX) models. ANNs are used as observers and pattern classifiers, and adaptive ARX models are used as observers. The generalized likelihood ratio (GLR) algorithm is used for change detection. Process faults are considered, and the robust FDI problem is also addressed. The approach is applied to a laboratory set-up tank system under closed-loop control.

Original language	English
Title of host publication	ETFA 2003
Subtitle of host publication	2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal
Place of Publication	Piscataway
Publisher	IEEE
Pages	739-746
Number of pages	8
Volume	2
ISBN (Print)	0-7803-7937-3
DOIs	https://doi.org/10.1109/ETFA.2003.1248772
Publication status	Published - 2003
Event	2003 IEEE Conference on Emerging Technologies and Factory Automation, ETFA 2003 - Lisbon, Portugal Duration: 16 Sep 2003 → 19 Sep 2003

Publication series

Name	IEEE International Conference on Emerging Technologies and Factory Automation, ETFA
ISSN (Print)	1946-0740

Conference

Conference	2003 IEEE Conference on Emerging Technologies and Factory Automation, ETFA 2003
Country/Territory	Portugal
City	Lisbon
Period	16/09/03 → 19/09/03

Keywords

Fault diagnosis
Fault detection
Neural networks
Robustness
Mathematical model
Power system modeling
Automatic control
Control systems
Manufacturing automation
Isolation technology

Access to Document

10.1109/ETFA.2003.1248772Licence: Unspecified

Cite this

Palma, L. B., Coito, F. V. D., & Silva, R. N. (2003). Fault diagnosis based on black-box models with application to a liquid-level system. In ETFA 2003: 2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal (Vol. 2, pp. 739-746). (IEEE International Conference on Emerging Technologies and Factory Automation, ETFA). IEEE. https://doi.org/10.1109/ETFA.2003.1248772

Palma, Luís Brito ; Coito, Fernando Vieira do ; Silva, Rui Neves. / Fault diagnosis based on black-box models with application to a liquid-level system. ETFA 2003: 2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal. Vol. 2 Piscataway : IEEE, 2003. pp. 739-746 (IEEE International Conference on Emerging Technologies and Factory Automation, ETFA).

@inproceedings{34e20dc2854342fdbbe8f867d315a9fa,

title = "Fault diagnosis based on black-box models with application to a liquid-level system",

abstract = "This paper proposes an on-line robust approach to fault detection and isolation (FDI) of dynamic systems. This FDI approach is based on black-box models: artificial neural networks (ANNs) and the autoregressive with exogenous input (ARX) models. ANNs are used as observers and pattern classifiers, and adaptive ARX models are used as observers. The generalized likelihood ratio (GLR) algorithm is used for change detection. Process faults are considered, and the robust FDI problem is also addressed. The approach is applied to a laboratory set-up tank system under closed-loop control.",

keywords = "Fault diagnosis, Fault detection, Neural networks, Robustness, Mathematical model, Power system modeling, Automatic control, Control systems, Manufacturing automation, Isolation technology",

author = "Palma, {Lu{\'i}s Brito} and Coito, {Fernando Vieira do} and Silva, {Rui Neves}",

year = "2003",

doi = "10.1109/ETFA.2003.1248772",

language = "English",

isbn = "0-7803-7937-3",

volume = "2",

series = "IEEE International Conference on Emerging Technologies and Factory Automation, ETFA",

publisher = "IEEE",

pages = "739--746",

booktitle = "ETFA 2003",

note = "2003 IEEE Conference on Emerging Technologies and Factory Automation, ETFA 2003 ; Conference date: 16-09-2003 Through 19-09-2003",

}

Palma, LB , Coito, FVD & Silva, RN 2003, Fault diagnosis based on black-box models with application to a liquid-level system. in ETFA 2003: 2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal. vol. 2, IEEE International Conference on Emerging Technologies and Factory Automation, ETFA, IEEE, Piscataway, pp. 739-746, 2003 IEEE Conference on Emerging Technologies and Factory Automation, ETFA 2003, Lisbon, Portugal, 16/09/03. https://doi.org/10.1109/ETFA.2003.1248772

Fault diagnosis based on black-box models with application to a liquid-level system. / Palma, Luís Brito ; Coito, Fernando Vieira do ; Silva, Rui Neves.

ETFA 2003: 2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal. Vol. 2 Piscataway : IEEE, 2003. p. 739-746 (IEEE International Conference on Emerging Technologies and Factory Automation, ETFA).

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

TY - GEN

T1 - Fault diagnosis based on black-box models with application to a liquid-level system

AU - Palma, Luís Brito

AU - Coito, Fernando Vieira do

AU - Silva, Rui Neves

PY - 2003

Y1 - 2003

N2 - This paper proposes an on-line robust approach to fault detection and isolation (FDI) of dynamic systems. This FDI approach is based on black-box models: artificial neural networks (ANNs) and the autoregressive with exogenous input (ARX) models. ANNs are used as observers and pattern classifiers, and adaptive ARX models are used as observers. The generalized likelihood ratio (GLR) algorithm is used for change detection. Process faults are considered, and the robust FDI problem is also addressed. The approach is applied to a laboratory set-up tank system under closed-loop control.

AB - This paper proposes an on-line robust approach to fault detection and isolation (FDI) of dynamic systems. This FDI approach is based on black-box models: artificial neural networks (ANNs) and the autoregressive with exogenous input (ARX) models. ANNs are used as observers and pattern classifiers, and adaptive ARX models are used as observers. The generalized likelihood ratio (GLR) algorithm is used for change detection. Process faults are considered, and the robust FDI problem is also addressed. The approach is applied to a laboratory set-up tank system under closed-loop control.

KW - Fault diagnosis

KW - Fault detection

KW - Neural networks

KW - Robustness

KW - Mathematical model

KW - Power system modeling

KW - Automatic control

KW - Control systems

KW - Manufacturing automation

KW - Isolation technology

UR - http://www.scopus.com/inward/record.url?scp=38149069208&partnerID=8YFLogxK

U2 - 10.1109/ETFA.2003.1248772

DO - 10.1109/ETFA.2003.1248772

M3 - Conference contribution

SN - 0-7803-7937-3

VL - 2

T3 - IEEE International Conference on Emerging Technologies and Factory Automation, ETFA

SP - 739

EP - 746

BT - ETFA 2003

PB - IEEE

CY - Piscataway

T2 - 2003 IEEE Conference on Emerging Technologies and Factory Automation, ETFA 2003

Y2 - 16 September 2003 through 19 September 2003

ER -

Palma LB , Coito FVD , Silva RN. Fault diagnosis based on black-box models with application to a liquid-level system. In ETFA 2003: 2003 IEEE Conference on Emerging Technologies and Factory Automation : proceedings : September 16-19, 2003, Lisbon, Portugal. Vol. 2. Piscataway: IEEE. 2003. p. 739-746. (IEEE International Conference on Emerging Technologies and Factory Automation, ETFA). https://doi.org/10.1109/ETFA.2003.1248772

Fault diagnosis based on black-box models with application to a liquid-level system

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this