Rolling bearing fault detection using a hybrid method based on Empirical Mode Decomposition and optimized wavelet multi-resolution analysis

In this paper, a hybrid method based on the combination of Empirical Mode Decomposition (EMD) and an optimized wavelet multi-resolution analysis (WMRA) is proposed. The pairing of these two time-frequency techniques is well adapted to analyze transient signals generated by rolling bearing defects. First, an optimal intrinsic mode function (IMF), having the most important kurtosis and covering the significant natural frequency, is selected using the classical EMD analysis. An envelope signal of the selected IMF's energy is calculated from Hilbert transform. This envelope is then analyzed by an optimized WMRA especially adapted to shock signals. A reconstructed signal is obtained and an envelope spectrum is performed to highlight the fault characteristic frequency. The results show that the proposed method can effectively get better time and frequency domain visualization of the fault occurrence compared to the application of WMRA or EMD alone.