Feature Extraction of Machine Vibration Using Lifting Wavelet Denoising and EMD and its Application in Fault Diagnosis

Abstract:

In order to reduce the random noise influence on empirical mode decomposition (EMD), the original data is adaptively denoised by lifting wavelet transform to strain mode aliasing, avoid the pseudo mode functions and improve the quality in EMD. The method is employed to analyze the rotor oil whirl vibration signal. Obtaining intrinsic mode functions (IMFs), the instantaneous frequency and amplitude can be calculated by Hilbert transform. Hilbert marginal spectrum can exactly provide the energy distribution of the signal with the change of instantaneous frequency. Thus, the characteristics information of the rotor oil whirl vibration signal can be extracted effectively. Experimental result demonstrate the validity of the proposed method.

• Periodical: Advanced Materials Research (Volumes 152-153)
• Pages: 383-386
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.152-153.383
• Online since: October 2010
• Authors: Feng Li Wang, Shu Lin Duan, Hong Tao Gao
• Keywords: Empirical Mode Decomposition (EMD), Fault Diagnosis, Feature Extraction, Lifting Wavelet

References

[1] Q.F. Meng，X. L. Jiang，and Z.J. He: Journal of Vibration, Measurement & Diagnosis, Vol. 17 (1997), p.25.

[2] N.E. Huang, Z. Shen and S.R. Long: Proceedings of the Royal Society of London, Series A454 (1998), p.903.

[3] W. Sweldens: Appl. Comput. Harmon. Anal. Vol, 3(1996) , p.186.

[4] R.L. Claypoole, M.D. Geoffrey, and W. Sweldens: IEEE Transactions on Image Processing, vol. 12 (2003), p.1449.

[5] G. J. Kong, P. L. Zhang, and J. J. Cao: Computer Engineering and Applications, vol. 44(2008), p.234.

[6] D. L. Donoho: IEEE Transaetions on Information Theory, vol. 41(1995), p.613.