Bearing Fault Diagnosis Using Gaussian Mixture Models (GMMs)


Article Preview

This paper presents a novel method for bearing fault diagnosis based on wavelet transform and Gaussian mixture models (GMMs). Vibration signals for normal bearings, bearings with inner race faults, outer race faults and ball faults were acquired from a motor-driven experimental system. The wavelet transform was used to process the vibration signals and to generate feature vectors. GMMs were trained and used as a diagnostic classifier. Experimental results have shown that GMMs can reliably classify different fault conditions and have a better classification performance as compared to the multilayer perceptron neural networks.



Edited by:

Kai Cheng, Yingxue Yao and Liang Zhou




J. Sun et al., "Bearing Fault Diagnosis Using Gaussian Mixture Models (GMMs)", Applied Mechanics and Materials, Vols. 10-12, pp. 553-557, 2008

Online since:

December 2007




[1] D.M. Yang, A.F. Stronach and P. MacConnel: Mechanical Systems and Signal Processing Vol. 16 (2002), pp.391-411.

[2] B. Samanta, Khamis R. Al-Balushi and Saeed A. Al-Araimi: Journal on Applied Signal Processing Vol. 3 (2004), pp.366-377.

[3] C.M. Alamo, F.J.C. Gil, C.T. Munilla and L.H. Gomez: Proceedings of IEEE International Conference of Acoustic Speech Signal Processing, (1996), pp.89-92.

[4] Ching-Tang Hsieh, Eugene Lai and Y.C. Wang: Journal of Information Science And Engineering Vol. 19 (2003), pp.267-282.

[5] Z.K. Peng and F.L. Chu: Mechanical Systems and Signal Processing Vol. 18 (2004), pp.199-221.

[6] S. Mallat: IEEE Transactions on Pattern Analysis and Machine Intelligence Vol. 2 (1989), pp.674-693.

[7] A. Dempster, N. Laird and D. Rubin: Journal of Royal Statistic Society, Vol. 39 (1977), pp.1-38.

[8] Information on http: /www. eecs. case. edu/laboratory/bearing/download. htm.