Application of the Hilbert – Huang Transform for Machine Fault Diagnostics

Abstract:

The vibration signals of the running machine contain non-stationary components. Usually, these non-stationary components contain abundant information on machine faults. In this paper, the Hilbert–Huang transform (HHT) method for the machine fault diagnosis is proposed. The empirical mode decomposition (EMD) method and Hilbert transform are key parts of the Hilbert–Huang transform method. The EMD will generate a collection of intrinsic mode functions (IMF). By applying EMD method and Hilbert transform to the vibration signal, we can get the Hilbert spectrum from which the faults in a running machine can be diagnosed and fault patterns can be identified. The practical vibration signals measured from roller machine with eccentric and friction faults are analysed by the Hilbert–Huang transform and Fourier transform in this paper. Finally, HHT’s performance in rolling machine fault detection is compared with that of the Fourier transform. The comparison results have shown that the HHT is superior than the Fourier transform in machine fault diagnostics. The different failure characteristic frequencies can be distinguished in the component of different orders of IMF, and the time and frequency of failure characteristic frequency appearance can be clearly reflected in the Hilbert spectrum.

• Periodical: Applied Mechanics and Materials (Volumes 182-183)
• Pages: 1484-1488
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.182-183.1484
• Online since: June 2012
• Authors: Zhao Yan Xuan, Miao Ge
• Keywords: Fault Diagnosis, Hilbert-Huang Transform (HHT), Non-Stationary Signals, Vibration Signal

References

[1] Z.K. Peng and F.L. Chus: Journal of Mechanical Systems and Signal Processing 18 (2004) p.199–221.

[2] N.E. Huang and others: Journal of Proc.R. Soc. Lond. A(1998).

[3] Dejin Yu, Weiping Lu and Robert G. Harrison: Physics Letters A , vol. 250 (1998), pp.323-327.

[4] B. Biswal , P.K. Dash b and B.K. Panigrahi : submitted to Journal of Applied Soft Computing (2009) p.107–117.

[5] Zhaoyan Xuan: The Prediction Study of the Atmosphere Time Sseries based on Empirical Mode Decomposition, Doctoral Dissertation(China University of Mining & Technology , Beijing ), In Chinese.

[6] Chunxiang Shi and Qifeng Luo: Journal of Acta Seismologica Sinca Vol. 16 No. 4 , pp.422-429.