Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method

Seyed Mojtaba Tabatabaeipour, Thomas Bak

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

Resumé

In this paper, a computationally efficient set-membership method for robust fault detection of linear systems is proposed. The method computes an interval outerapproximation of the output of the system that is consistent with the model, the bounds on noise and disturbance, and the past measurements. If the output of the system does not belong to this interval, a fault is detected. To compute the output interval, we propose using support functions. Only two support functions for each output must be computed which results in a computationally efficient algorithm. Moreover, the method is trivially parallelizable. The method is demonstrated for fault detection of a hydraulic pitch actuator of a wind turbine. We show the effectiveness of the proposed method by comparing our results with two zonotope-based set-membership methods.

Published in:
OriginalsprogEngelsk
TitelControl Applications (CCA), 2014 IEEE Conference on
Antal sider6
ForlagIEEE Press
Publikationsdato2014
Sider358-363
ISBN (Trykt)978-1-4799-7408-5
DOI
StatusUdgivet - 2014
Begivenhed2014 IEEE Conference on Control Applications - Juan Les Antibes, Frankrig
Varighed: 8 okt. 201410 okt. 2014
Konferencens nummer: 35444

Konference

Konference2014 IEEE Conference on Control Applications
Nummer35444
LandFrankrig
ByJuan Les Antibes
Periode08/10/201410/10/2014
NavnI E E E International Conference on Control Applications. Proceedings
ISSN1085-1992

Fingerprint

Fault detection
Linear systems
Hydraulic actuators
Wind turbines

Citer dette

Tabatabaeipour, S. M., & Bak, T. (2014). Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method. I Control Applications (CCA), 2014 IEEE Conference on (s. 358-363). IEEE Press. I E E E International Conference on Control Applications. Proceedings https://doi.org/10.1109/CCA.2014.6981372
Tabatabaeipour, Seyed Mojtaba ; Bak, Thomas. / Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method. Control Applications (CCA), 2014 IEEE Conference on. IEEE Press, 2014. s. 358-363 (I E E E International Conference on Control Applications. Proceedings).
@inproceedings{e6a64eedfe6a463d9c5403c55d83d50b,
title = "Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method",
abstract = "In this paper, a new computationally efficientset-membership method for robust fault detection of linearsystems is proposed. The method computes an intervalouter-approximation of the output of the system that isconsistent with the model, the bounds on noise anddisturbance, and the past measurements. If the output ofthe system does not belong to this interval, a fault isdetected. To compute the output interval, we propose usingsupport functions. Only two support functions for eachoutput must be computed which results in a computationallyefficient algorithm. Moreover, the method is triviallyparallelizable. The method is demonstrated for faultdetection of a hydraulic pitch actuator of a wind turbine.We show the effectiveness of the proposed method bycomparing our results with two zonotope-basedset-membership methods.",
author = "Tabatabaeipour, {Seyed Mojtaba} and Thomas Bak",
year = "2014",
doi = "10.1109/CCA.2014.6981372",
language = "English",
isbn = "978-1-4799-7408-5",
pages = "358--363",
booktitle = "Control Applications (CCA), 2014 IEEE Conference on",
publisher = "IEEE Press",

}

Tabatabaeipour, SM & Bak, T 2014, Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method. i Control Applications (CCA), 2014 IEEE Conference on. IEEE Press, I E E E International Conference on Control Applications. Proceedings, s. 358-363, Juan Les Antibes, Frankrig, 08/10/2014. https://doi.org/10.1109/CCA.2014.6981372

Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method. / Tabatabaeipour, Seyed Mojtaba; Bak, Thomas.

Control Applications (CCA), 2014 IEEE Conference on. IEEE Press, 2014. s. 358-363.

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method

AU - Tabatabaeipour, Seyed Mojtaba

AU - Bak, Thomas

PY - 2014

Y1 - 2014

N2 - In this paper, a new computationally efficientset-membership method for robust fault detection of linearsystems is proposed. The method computes an intervalouter-approximation of the output of the system that isconsistent with the model, the bounds on noise anddisturbance, and the past measurements. If the output ofthe system does not belong to this interval, a fault isdetected. To compute the output interval, we propose usingsupport functions. Only two support functions for eachoutput must be computed which results in a computationallyefficient algorithm. Moreover, the method is triviallyparallelizable. The method is demonstrated for faultdetection of a hydraulic pitch actuator of a wind turbine.We show the effectiveness of the proposed method bycomparing our results with two zonotope-basedset-membership methods.

AB - In this paper, a new computationally efficientset-membership method for robust fault detection of linearsystems is proposed. The method computes an intervalouter-approximation of the output of the system that isconsistent with the model, the bounds on noise anddisturbance, and the past measurements. If the output ofthe system does not belong to this interval, a fault isdetected. To compute the output interval, we propose usingsupport functions. Only two support functions for eachoutput must be computed which results in a computationallyefficient algorithm. Moreover, the method is triviallyparallelizable. The method is demonstrated for faultdetection of a hydraulic pitch actuator of a wind turbine.We show the effectiveness of the proposed method bycomparing our results with two zonotope-basedset-membership methods.

U2 - 10.1109/CCA.2014.6981372

DO - 10.1109/CCA.2014.6981372

M3 - Article in proceeding

SN - 978-1-4799-7408-5

SP - 358

EP - 363

BT - Control Applications (CCA), 2014 IEEE Conference on

PB - IEEE Press

ER -

Tabatabaeipour SM, Bak T. Robust Fault Detection of Linear Systems using a Computationally Efficient Set-membership Method. I Control Applications (CCA), 2014 IEEE Conference on. IEEE Press. 2014. s. 358-363. (I E E E International Conference on Control Applications. Proceedings). https://doi.org/10.1109/CCA.2014.6981372