LAUSR

A rolling bearing vibration signal fault feature enhancement method based on adaptive complete ensemble empirical mode decomposition with adaptive noise algorithm (CEEMDAN) and maximum correlated kurtosis deconvolution (MCKD) is proposed to address the issue that rolling bearings are prone to noise in the early stage and difficult to extract feature information accurately. The method uses the CEEMDAN algorithm to reduce the noise of the rolling bearing vibration signal in the first step; then, the MCKD algorithm is used to deconvolve the signal to enhance the weak shock components in the signal and improve the SNR. Finally, the envelope spectrum analysis is performed to extract the feature frequencies. Simulation and experimental results show that the CEEMDAN-MCKD method can highlight the fault characteristic frequency and multiplier frequency better than other methods and realize the characteristic enhancement of incipient fault vibration signals of rolling bearings under constant and variable operating conditions.