An estimating method for initial faulty time of rolling element bearing based on EEMD and spectral kurtosis

Envelope demodulation analysis is an effective method to deal with vibration signals captured from rolling bearings. However, it is not easy to separate the fault component from the noise because the fault signals in presence of strong masking noise that happens in the early stage of bearing failure. Therefore, an approach to predict the initial failure time of rolling bearing based on the combination of EEMD and spectral kurtosis is proposed in this paper. On the one hand, EEMD is used to weaken the low-frequency interference and highlight the high-frequency resonance component. On the other hand, spectral kurtosis (SK) has a great ability of diagnostic detecting. The integration of the two methods can give the energy change of filtered signal over time distinctly by analyzing the historical data. Then the initial failure time of rolling bearing can be estimated. The method proposed in this paper is verified through an engineering example signal. The results show that this method can identify the faults earlier.