Empirical Mode Decomposition Combined with Empirical Wavelets for Extracting Bearing Frequencies in a Noisy Environment and Early Detection of Defects

The amplitude demodulation of a bearing signal allows for the extraction of component information-carrying defects on rotary machines. However, the quality of the demodulated signal depends on the selected frequency band for demodulation. Kurtogram is widely used to detect the frequency bandwidth which is the most excited by a defect. However in presence of high noises, the Kurtogram may be deficient in effectively detecting the resonances and it presents some instabilities. In the last decade, the Empirical Mode Decomposition (EMD) technique has been used by a lot of researchers for the signal decomposition. In this study, the EMD and Empirical Wavelet (EW) are used to generate a new feature. The EW is used to generate a filter bank which depends on the content of the component frequencies of the signal. A segmentation of the spectrum to define the support boundaries of the filter is proposed. The new indicator is proposed in order to track the frequency band that is more excited by a bearing fault. This study shows that the proposed technique can detect the resonances in all cases of simulation. On the other hand, the proposed method is able first to detect the resonance frequencies and secondly to detect on which Intrinsic Mode Function (IMF), the bearing default occurs. The proposed technique has confirmed its effectiveness by testing it on experimental signals obtained from a test bench with defects on a bearing outer race. A defect of only 40 μ on the outer race has been detected, which makes this method very effective for an early detection of bearing defects.

[1]  Ioannis Antoniadis,et al.  Demodulation of Vibration Signals Generated by Defects in Rolling Element Bearings Using Complex Shifted Morlet Wavelets , 2002 .

[2]  Joseph Mathew,et al.  Multiple Band-Pass Autoregressive Demodulation for Rolling-Element Bearing Fault Diagnosis , 2001 .

[3]  Jérôme Gilles,et al.  Empirical Wavelet Transform , 2013, IEEE Transactions on Signal Processing.

[4]  Robert B. Randall,et al.  Rolling element bearing fault diagnosis based on the combination of genetic algorithms and fast kurtogram , 2009 .

[5]  Yuh-Tay Sheen,et al.  A complex filter for vibration signal demodulation in bearing defect diagnosis , 2004 .

[6]  Steven A. Orszag,et al.  CBMS-NSF REGIONAL CONFERENCE SERIES IN APPLIED MATHEMATICS , 1978 .

[7]  Bin Chen,et al.  Defect Detection for Wheel-Bearings with Time-Spectral Kurtosis and Entropy , 2014, Entropy.

[8]  N. Tandon,et al.  A review of vibration and acoustic measurement methods for the detection of defects in rolling element bearings , 1999 .

[9]  Robert B. Randall,et al.  The spectral kurtosis: application to the vibratory surveillance and diagnostics of rotating machines , 2006 .

[10]  Reza Ansari,et al.  Prediction of Vibrational Behavior of Grid-Stiffened Cylindrical Shells , 2014 .

[11]  Antoine Tahan,et al.  Monitoring Machines by Using a Hybrid Method Combining MED, EMD, and TKEO , 2014 .

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

[13]  Ingrid Daubechies,et al.  Ten Lectures on Wavelets , 1992 .

[14]  Robert X. Gao,et al.  Mechanical Systems and Signal Processing Approximate Entropy as a Diagnostic Tool for Machine Health Monitoring , 2006 .

[15]  Hai Qiu,et al.  Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics , 2006 .

[16]  Hong Fan,et al.  Rotating machine fault diagnosis using empirical mode decomposition , 2008 .

[17]  Ming J. Zuo,et al.  GEARBOX FAULT DIAGNOSIS USING ADAPTIVE WAVELET FILTER , 2003 .

[18]  Tomasz Barszcz,et al.  A novel method for the optimal band selection for vibration signal demodulation and comparison with the Kurtogram , 2011 .