Fault feature extraction based on combination of envelope order tracking and cICA for rolling element bearings

Abstract Vibration from incipient faults of rolling element bearings (REBs) is usually too weak to be observed in a conventional spectrum analysis. The envelope analysis or high-frequency resonance technique is an effective tool for the incipient fault detection of REBs. The newly developed envelope order tracking, an improved version of the envelope analysis, can be well performed even in a varying-speed condition. However, the envelope order tracking can be invalid for multi-impulsive sources. To address this issue, a scheme for the weak feature extraction of faulty REBs has been proposed in this paper by combining the envelope order tracking and the constrained independent component analysis (cICA). In the proposed scheme, the envelope order tracking approach is utilized to obtain the envelopes of sensor observed mixtures at different positions. Then, the envelopes are turned from the time domain into the angle domain by the constant-angle increments resampling scheme in the computed order tracking (COT). Subsequently, the cICA method is employed to extract the interesting envelope independent components (ICs) by a reference signal, which is constructed according to the prior-known feature frequency of the bearing. As a result, the faults related features can be clearly exposed in the spectra of the obtained interesting envelope ICs. Simulations and experimental results support the proposed method positively.

[1]  Zhi-Lin Zhang,et al.  Morphologically constrained ICA for extracting weak temporally correlated signals , 2008, Neurocomputing.

[2]  Jing Na,et al.  Envelope order tracking for fault detection in rolling element bearings , 2012 .

[3]  K. R. Fyfe,et al.  ANALYSIS OF COMPUTED ORDER TRACKING , 1997 .

[4]  Yu Zhou,et al.  Constrained independent component analysis and its application to machine fault diagnosis , 2011 .

[5]  Swathi Kavuri Sri,et al.  Extracting EEG rhythms using ICA-R , 2008, 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence).

[6]  Wei Lu,et al.  ICA with Reference , 2006, Neurocomputing.

[7]  Lei Hu,et al.  A novel manifold learning denoising method on bearing vibration signals , 2016 .

[8]  Chen Lu,et al.  An intelligent approach to machine component health prognostics by utilizing only truncated histories , 2014 .

[9]  Kok Kiong Tan,et al.  Order-crossing removal in Gabor order tracking by independent component analysis , 2009 .

[10]  Erkki Oja,et al.  Independent component analysis: algorithms and applications , 2000, Neural Networks.

[11]  P. D. McFadden,et al.  Vibration monitoring of rolling element bearings by the high-frequency resonance technique — a review , 1984 .

[12]  Min-Chun Pan,et al.  Investigation on improved Gabor order tracking technique and its applications , 2006 .

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

[14]  Min-Chun Pan,et al.  Adaptive Vold Kalman filtering order tracking , 2007 .

[15]  Jian Ma,et al.  Health assessment of rotary machinery based on integrated feature selection and Gaussian mixed model , 2014 .

[16]  Kok Kiong Tan,et al.  High efficient crossing-order decoupling in Vold-Kalman filtering order tracking based on independent component analysis , 2010 .

[17]  Guillaume Gelle,et al.  BLIND SOURCE SEPARATION: A TOOL FOR ROTATING MACHINE MONITORING BY VIBRATIONS ANALYSIS? , 2001 .

[18]  Robert B. Randall,et al.  Single and multi-stage phase demodulation based order-tracking , 2014 .

[19]  Jing Na,et al.  Envelope extraction based dimension reduction for independent component analysis in fault diagnosis of rolling element bearing , 2014 .

[20]  Chen Lu,et al.  Bearing performance degradation assessment and prediction based on EMD and PCA-SOM , 2014 .

[21]  Jianshe Kang,et al.  Bearing fault diagnosis and degradation analysis based on improved empirical mode decomposition and maximum correlated kurtosis deconvolution , 2015 .

[22]  Robert B. Randall,et al.  Rolling element bearing diagnostics—A tutorial , 2011 .

[23]  F. Bonnardot,et al.  Impact of the non-uniform angular sampling on mechanical signals , 2014 .