Fault detection and diagnosis using an art-based neural network

Keem Siah Yap, Mau Teng Au, Chee Peng Lim, Junita Mohamad Saleh

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

Abstract

The Generalized Adaptive Resonance Theory (GART) network is a neural network model based on the integration of Gaussian ARTMAP and the Generalized Regression Neural Network. It is capable of online learning, and is effective in tackling classification as well as regression tasks, as demonstrated in our previous work. In this paper, we further enhance the capability of the GART network with the Laplacian functions and with new vigilance and match-tracking mechanisms. In addition, a rule extraction procedure is incorporated into its dynamics, and its applicability to fault detection and diagnosis tasks is assessed. IF-THEN rules can be extracted from the weights of the trained GART network after a pruning process. The classification and rule extraction capability of GART are evaluated using one benchmark data set from medical application, and one real data set collected from a power generation plant. These results are then compared with those reported by other methods. The outcomes demonstrate that GART is able to produce high classification rates with quality rules for tackling fault detection and diagnosis problems.

Original languageEnglish
Title of host publicationProceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010
Pages118-125
Number of pages8
Publication statusPublished - 20 Jul 2010
Event10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010 - Innsbruck, Austria
Duration: 15 Feb 201017 Feb 2010

Publication series

NameProceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010

Other

Other10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010
CountryAustria
CityInnsbruck
Period15/02/1017/02/10

Fingerprint

Fault detection
Failure analysis
Neural networks
Circuit theory
Medical applications
Power generation

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Science Applications

Cite this

Yap, K. S., Au, M. T., Lim, C. P., & Saleh, J. M. (2010). Fault detection and diagnosis using an art-based neural network. In Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010 (pp. 118-125). (Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010).
Yap, Keem Siah ; Au, Mau Teng ; Lim, Chee Peng ; Saleh, Junita Mohamad. / Fault detection and diagnosis using an art-based neural network. Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010. 2010. pp. 118-125 (Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010).
@inproceedings{a13e88221a504c9c97eadca04da25278,
title = "Fault detection and diagnosis using an art-based neural network",
abstract = "The Generalized Adaptive Resonance Theory (GART) network is a neural network model based on the integration of Gaussian ARTMAP and the Generalized Regression Neural Network. It is capable of online learning, and is effective in tackling classification as well as regression tasks, as demonstrated in our previous work. In this paper, we further enhance the capability of the GART network with the Laplacian functions and with new vigilance and match-tracking mechanisms. In addition, a rule extraction procedure is incorporated into its dynamics, and its applicability to fault detection and diagnosis tasks is assessed. IF-THEN rules can be extracted from the weights of the trained GART network after a pruning process. The classification and rule extraction capability of GART are evaluated using one benchmark data set from medical application, and one real data set collected from a power generation plant. These results are then compared with those reported by other methods. The outcomes demonstrate that GART is able to produce high classification rates with quality rules for tackling fault detection and diagnosis problems.",
author = "Yap, {Keem Siah} and Au, {Mau Teng} and Lim, {Chee Peng} and Saleh, {Junita Mohamad}",
year = "2010",
month = "7",
day = "20",
language = "English",
isbn = "9780889868182",
series = "Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010",
pages = "118--125",
booktitle = "Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010",

}

Yap, KS, Au, MT, Lim, CP & Saleh, JM 2010, Fault detection and diagnosis using an art-based neural network. in Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010. Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010, pp. 118-125, 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010, Innsbruck, Austria, 15/02/10.

Fault detection and diagnosis using an art-based neural network. / Yap, Keem Siah; Au, Mau Teng; Lim, Chee Peng; Saleh, Junita Mohamad.

Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010. 2010. p. 118-125 (Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010).

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

TY - GEN

T1 - Fault detection and diagnosis using an art-based neural network

AU - Yap, Keem Siah

AU - Au, Mau Teng

AU - Lim, Chee Peng

AU - Saleh, Junita Mohamad

PY - 2010/7/20

Y1 - 2010/7/20

N2 - The Generalized Adaptive Resonance Theory (GART) network is a neural network model based on the integration of Gaussian ARTMAP and the Generalized Regression Neural Network. It is capable of online learning, and is effective in tackling classification as well as regression tasks, as demonstrated in our previous work. In this paper, we further enhance the capability of the GART network with the Laplacian functions and with new vigilance and match-tracking mechanisms. In addition, a rule extraction procedure is incorporated into its dynamics, and its applicability to fault detection and diagnosis tasks is assessed. IF-THEN rules can be extracted from the weights of the trained GART network after a pruning process. The classification and rule extraction capability of GART are evaluated using one benchmark data set from medical application, and one real data set collected from a power generation plant. These results are then compared with those reported by other methods. The outcomes demonstrate that GART is able to produce high classification rates with quality rules for tackling fault detection and diagnosis problems.

AB - The Generalized Adaptive Resonance Theory (GART) network is a neural network model based on the integration of Gaussian ARTMAP and the Generalized Regression Neural Network. It is capable of online learning, and is effective in tackling classification as well as regression tasks, as demonstrated in our previous work. In this paper, we further enhance the capability of the GART network with the Laplacian functions and with new vigilance and match-tracking mechanisms. In addition, a rule extraction procedure is incorporated into its dynamics, and its applicability to fault detection and diagnosis tasks is assessed. IF-THEN rules can be extracted from the weights of the trained GART network after a pruning process. The classification and rule extraction capability of GART are evaluated using one benchmark data set from medical application, and one real data set collected from a power generation plant. These results are then compared with those reported by other methods. The outcomes demonstrate that GART is able to produce high classification rates with quality rules for tackling fault detection and diagnosis problems.

UR - http://www.scopus.com/inward/record.url?scp=77954581394&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954581394&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9780889868182

T3 - Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010

SP - 118

EP - 125

BT - Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010

ER -

Yap KS, Au MT, Lim CP, Saleh JM. Fault detection and diagnosis using an art-based neural network. In Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010. 2010. p. 118-125. (Proceedings of the 10th IASTED International Conference on Artificial Intelligence and Applications, AIA 2010).