Factorial Hidden Markov Model Recognition Method Based on Multi-Channel Information Fusion

Abstract:

Combining Self-organizing Feature Map (SOM) and Factorial hidden Markov model (FHMM), a new FHMM fault recognition method based on multi-sensor vibration information fusion is proposed. In the proposed method, the SOM neural network is used to reduce the information redundancy in feature vectors extracted from the multi-sensor’s vibration measurements, FHMM as a classifier. The fault recognition in the speed-up and speed-down process of rotating machinery was successfully completed. The experiment result shows that the proposed method is very effective.

• Periodical: Advanced Materials Research (Volumes 291-294)
• Pages: 2027-2033
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.291-294.2027
• Online since: July 2011
• Authors: Zhi Nong Li, Guan Hua Wu, Jing Jiang, Fu Zhou Feng
• Keywords: Factorial Hidden Markov Model, Fault Diagnosis, Information Fusion, Redundancy Reduction, Self-Organizing Feature Map

References

[1] Feng C.J. Hidden Markov Models dynamic pattern recognition theories, methods and its applications in faults diagnosis of rotating machine [Doctoral Dissertation]. Zhejiang University, China, 2002, in Chinese.

[2] Ghahramani Z., Jordan M., Factorial hidden Markov models. Machine Learning, 1997, 29(3): 245～275.

[3] Logan B, Moreno P. Factorial HMMs for acoustic modeling. IEEE Transactions on Acoustic,Speech and Signal Processing, 1998, 2(2): 813～816.

DOI: 10.1109/icassp.1998.675389

[4] Li Z.N., Wu Z.T., He Y.Y, Chu F.L. Hidden Markov Model-based fault diagnostics method in speed-up and speed-down process for rotating machinery. Mechanical System and Signal Processing, 2005, 19(1-2): 329-339.

DOI: 10.1016/j.ymssp.2004.01.001

[5] Li Z.N., Ding Q.Q., Wu Z.T., Feng C.J. Study on bispectrum-fhmm recognition method in speed-up and speed-down process of rotating machinery. Journal of Vibration Engineering, 2003, 16(2): 171~174, in Chinese.

[6] Li Z.N, Wu Z.T., Ding Q.Q, He Y.Y., Chu F.L. Application of wavelet transform and fhmm in fault diagnosis of speed-up and speed-down process of rotating machinery. Control Engineering of China, 2003, 10(4): 299~302, in Chinese.

[7] Li Z.N, Ding Q.Q., Feng C.J., Wu Z.T. Study on BSI-FHMM recognition method in speed-up and speed-down process of rotating machinery. China Mechanical Engineering, 2002, 13(20): 1730~1733, in Chinese.

[8] Li Z.N, Zeng M.R., Han J., He Y.Y, Chu F.L. Application of principal component analysis and factorial hidden markov model in machine fault diagnosis. Journal of Mechanical Strength, 2007, 29(1): 25~29, in Chinese.

[9] Li Z.N, Hao W., Han J., Chu F.L., Wu Z.T. Factorial hidden markov model recognition method based on blind identification of nonlinear time series models. Chinese Journal of Mechanical Engineering, 2007, 43(1):191~195, in Chinese.

DOI: 10.3901/jme.2007.01.191

[10] Ding Q.Q., Li Z.N, Wu Z.T., Zheng S.X. Research on fault diagnosis method for rotating machine based on factorial hidden markov model. Power Engineering, 2003, 23(4): 2560~2563, in Chinese.