The Fault Diagnosis Method Based on Wavelet Packet Characteristic Entropy and Relevance Vector Machine

Abstract:

The sensor fault diagnosis method based on wavelet packet characteristic entropy and relevance vector machine are researched, which is used on engine test bed . In detail, the wavelet packet characteristic entropy is applied for feature extraction to get the feature matrix which denote all kinds of known working status of sensor, and the feature matrix are encoded; The feature matrix as inputs and feature codes as outputs are proposed for training the relevance vector machine classifier to get the optimum parameters. In the fault diagnosis unit, it uses the wavelet packet characteristic entropy to acquire the on-line feature matrix of sensor. The on-line feature matrix is supplied to the trained relevance vector machine classifier as inputs to validate the working status of sensor. An application of the method in engine test bed system is introduced. Finally, the applicability and effectiveness of the method is illustrated by experiments

• Periodical: Applied Mechanics and Materials (Volume 151)
• Pages: 3-6
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.151.3
• Online since: January 2012
• Authors: Shu Yan Zhao, Qi Wang
• Keywords: Engine Test Bed, Relevance Vector Machine, Sensor Fault Diagnosis, Wavelet Packet Characteristic Entropy

References

[1] GUI Zhong-hua, HAN Feng-qin, Neural network based on wavelet packet-charicteristic entropy for fault diagnosis of draft tube, Proceedings of the CSEE, 25(2005), 99-102.

[2] Tipping M E, Sparse Bayesian learning and the relevance vector machine , Journal of Machine Learning Research, 2001: 211-244.

[3] Silva C, Ribeiro B , Scaling Tex t Classification with Relevance Vector Machines . IEEE International Conference on System s, Man and Cybernetics. 2006: 4186 -4191.

DOI: 10.1109/icsmc.2006.384791

[4] Wahyu C, Achmad W, Y BS, Application of relevance vector machine and logistic regression for machine degradation assessment, Mechanical Systems and Signal Processing. 2010, 1161-1171.

DOI: 10.1016/j.ymssp.2009.10.011

[5] Dimitris G. Tzikas, Aristidis C. Likas, Sparse Bayesian Modeling With Adaptive Kernel Learning, IEEE TRANSACTIONS ON NEURAL NETWORKS, 20(2009), 926-937.

DOI: 10.1109/tnn.2009.2014060

[6] Jin Yuana, Kesheng Wang, Integrating relevance vector machines and genetic algorithms for optimization of seed-separating process, Engineering Applications of Artificial Intelligence 20 (2007) 970–979.

DOI: 10.1016/j.engappai.2007.01.002