Bearing fault diagnosis based on redundant second generation wavelet denoising and EEMD

Ensemble empirical mode decomposition (EEMD) is a newly noise-assisted analysis method, is proposed to eliminating mode mixing present in the original EMD, which has been proved quite effective in the bearing fault diagnosis. But when the Signal-to-Noise(SNR) of the vibration signal is low, it hardly extracts the fault information. Worse more, mode mixing appear frequently when EMD is applied to the vibration signals collected in real life. To improve the accuracy of bearing fault diagnosis, a new method based on redundant second generation wavelet denoising and EEMD is proposed in this paper. Simulated signals demonstrate the effectiveness of the new method in extracting characteristic frequency from a vibration signal with low SNR. Finally, the new method is applied to diagnosing a bearing with fault, and the result proves that the new method is effective in bearing fault diagnosis.