Development of a Method for Selection of Effective Singular Values in Bearing Fault Signal De-Noising

Singular value decomposition (SVD) is a widely used and powerful tool for signal extraction under noise. Noise attenuation relies on the selection of the effective singular value because these values are significant features of the useful signal. Traditional methods of selecting effective singular values (or selecting the useful components to rebuild the faulty signal) consist of seeking the maximum peak of the differential spectrum of singular values. However, owing to the small number of selected effective singular values, these methods lead to excessive de-noised effects. In order to get a more appropriate number of effective singular values, which preserves the components of the original signal as much as possible, this paper used a difference curvature spectrum of incremental singular entropy to determine the number of effective singular values. Then the position was found where the difference of two peaks in the spectrum declines in an infinitely large degree for the first time, and this position was regarded as the boundary of singular values between noise and a useful signal. The experimental results showed that the modified methods could accurately extract the non-stationary bearing faulty signal under real background noise.

[1]  Taiyong Wang,et al.  On the applications of SVD in fault diagnosis , 2003, SMC'03 Conference Proceedings. 2003 IEEE International Conference on Systems, Man and Cybernetics. Conference Theme - System Security and Assurance (Cat. No.03CH37483).

[2]  Xuezhi Zhao,et al.  Selection of effective singular values using difference spectrum and its application to fault diagnosis of headstock , 2011 .

[3]  Robert B. Randall.  Frequency Analysis , 1987 .

[4]  Tao Xie,et al.  Fault Diagnosis of Analog Circuit Based on High-Order Cumulants and Information Fusion , 2014, J. Electron. Test..

[5]  Danni Ai,et al.  Adaptive Tensor-Based Principal Component Analysis for Low-Dose CT Image Denoising , 2015, PloS one.

[6]  David L. Donoho,et al.  Optimal Shrinkage of Singular Values , 2014, IEEE Transactions on Information Theory.

[7]  Pradeep K. Atrey,et al.  Analysis and extension of multiresolution singular value decomposition , 2014, Inf. Sci..

[8]  Gang Chen,et al.  Study on Hankel matrix-based SVD and its application in rolling element bearing fault diagnosis , 2015 .

[9]  Robert B. Randall,et al.  A Stochastic Model for Simulation and Diagnostics of Rolling Element Bearings With Localized Faults , 2003 .

[10]  R. Tandra,et al.  Fundamental limits on detection in low SNR under noise uncertainty , 2005, 2005 International Conference on Wireless Networks, Communications and Mobile Computing.

[11]  Chun-Chieh Wang,et al.  Feature Extraction Techniques of Non-Stationary Signals for Fault Diagnosis in Machinery Systems , 2012 .

[12]  Yi-Ping Hung,et al.  Single-Pass K-SVD for Efficient Dictionary Learning , 2014, Circuits Syst. Signal Process..

[13]  Michael G. Pecht,et al.  Rolling element bearing fault diagnosis using simulated annealing optimized spectral kurtosis , 2013, 2013 IEEE Conference on Prognostics and Health Management (PHM).

[14]  Jong-Beom Lee,et al.  A validated accurate fault location approach by applying noise cancellation technique , 2012 .

[15]  Pei Lin Zhang,et al.  Bearing Fault Detection Based on SVD and EMD , 2012 .

[16]  Jing Tian,et al.  Motor Bearing Fault Detection Using Spectral Kurtosis-Based Feature Extraction Coupled With K-Nearest Neighbor Distance Analysis , 2016, IEEE Transactions on Industrial Electronics.

[17]  Nikhil R. Pal,et al.  Feature selection with SVD entropy: Some modification and extension , 2014, Inf. Sci..

[18]  Gabriela Achtenova,et al.  Modern Vibration Signal Processing Techniques for Vehicle Gearbox Fault Diagnosis , 2014 .

[19]  P. McFadden.  Examination of a technique for the early detection of failure in gears by signal processing of the time domain average of the meshing vibration , 1987 .

[20]  Nagi Gebraeel,et al.  A Neural Network Degradation Model for Computing and Updating Residual Life Distributions , 2008, IEEE Transactions on Automation Science and Engineering.

[21]  Sunil K Jha,et al.  Denoising by Singular Value Decomposition and Its Application to Electronic Nose Data Processing , 2011, IEEE Sensors Journal.

[22]  J. Antoni.  Fast computation of the kurtogram for the detection of transient faults , 2007 .

[23]  Andrew B. Nobel,et al.  Reconstruction of a low-rank matrix in the presence of Gaussian noise , 2010, J. Multivar. Anal..

[24]  Baojun Zhao,et al.  A Perceptual Image Quality Assessment Metric Using Singular Value Decomposition , 2015, Circuits Syst. Signal Process..

[25]  Guanghua Xu,et al.  Detection of weak transient signals based on wavelet packet transform and manifold learning for rolling element bearing fault diagnosis , 2015 .

[26]  P. Landa.  Mechanism of stochastic resonance , 2004 .