Fault Diagnosis of Rotating Machinery Base on Wavelet Packet Energy Moment and HMM

Chun Liang Zhang; Xia Yue; Sheng Li; Jian Li

doi:10.4028/www.scientific.net/KEM.455.558

Paper Titles

On-Line Prediction Model of Ultrasonic Polishing Surface Roughness
p.539

Analysis on Forming Characteristics of Cup-Shaped Thin-Walled Inner Gear Spinning
p.544

A Study on Chip Breakage in Mini-Pore Vibration Drilling for Hard-to-Cut Material of Austenitic Stainless Steel
p.548

Intelligent Design of a Mechanical Part Library in Real-Time Collaborative Design
p.553

Fault Diagnosis of Rotating Machinery Base on Wavelet Packet Energy Moment and HMM
p.558

Profile Error Synthetic Prediction Model for One-Side Milling of Spatial Cams
p.565

Research on Grinding Technology of Isometric Polygonal Profile
p.571

Research on the Unstability Motion of BTA Deep Hole Boring Bar
p.575

Influence of Original Structure on the Grind-Hardened Layer Structure and its Property of 65Mn Steel
p.580

HomeKey Engineering MaterialsKey Engineering Materials Vol. 455Fault Diagnosis of Rotating Machinery Base on...

Fault Diagnosis of Rotating Machinery Base on Wavelet Packet Energy Moment and HMM

Abstract:

A feature extraction method is presented for the fault diagnosis of large rotating machinery to improve the performance of on-line monitoring. According to the characteristics of fault vibration signals, wavelet packet decomposition, a well-known tool to multi-scale analysis, is applied to extracting frequency band energy features; and Hidden Markova Model (HMM) is used to classify. The final feature array is composed of time-domain, amplitude-domain features and wavelet packet energy moment features which reflect the inherent energy distribution characteristics of different faults. The continuous density hidden Markov model (CDHMM) is adopted to recognize the state in on-line monitoring, and the diagnosis success rates are more than 91% to six typical faults on different rotation speeds. The experimental results show the fault diagnosis system is valid and robust, particularly the method of feature extraction.

You might also be interested in these eBooks

Manufacturing Automation Technology Development

View Preview

Info:

Periodical:

Key Engineering Materials (Volume 455)

Pages:

558-564

DOI:

https://doi.org/10.4028/www.scientific.net/KEM.455.558

Citation:

Cite this paper

Online since:

December 2010

Authors:

Chun Liang Zhang, Xia Yue, Sheng Li, Jian Li

Keywords:

Fault Diagnosis, Hidden Markov Model (HMM), Wavelet Packet Energy Moment

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

References

[1] S. Z. Shun and Q. L. Yin: Technology and Application of Condition Monitoring and Fault Diagnosis. (Chemical Industry Press, China 2003), pp.38-42 (In Chinese).

Google Scholar

[2] R.T. Hou. Research on Fault Diagnosis Based on the Complexity of Modern Nonlinear Theory (Ph.D., Northeastern University, China, 2004).

Google Scholar

[3] H. B. Zheng. Study on Non-stationary Gaussian Signal Feature Extraction and Fault Diagnosis(Ph.D., Hefei University of Technology, China, 2003).

Google Scholar

[4] Q. B. Zhu. Research on Feature Extraction of non-stationary signal based on Wavelet Theory and Intelligent Diagnosis(Ph.D., Northeastern University, China, 2005).

Google Scholar

[5] J. M. Lee, S. J. Kim and Y. Song. Journal of Sound and Vibration, Vol. 276 (2004) No. 3-5, pp.1065-1080.

Google Scholar

[6] V. Purushotham, S. Narayanan and Suryanarayana A.N. Prasad. NDT&E International, Vol. 38 (2005) No. 8, pp.654-664.

Google Scholar

[7] P. L. Zhao, Z. M. Xi and H. Xiao. Radar Science and Technology, Vol. 5 (2007) No. 4, pp.242-245.

Google Scholar