Adaptive generalized empirical wavelet transform and its application to fault diagnosis of rolling bearing

This paper proposes a new signal decomposition method named adaptive generalized empirical wavelet transform (AGEWT) to address the limitation of the decomposition effect of empirical wavelet transform (EWT) on non-stationary signals caused by the number of initial decomposition modes. AGEWT method first introduces the amplitude distribution spectrum (ADS) to determine the spectral segmentation boundaries for eliminating the need of dependence on the number of decomposition modes. Subsequently, generalized filter is defined to reconstruct each frequency band signals to effectively suppress the noises in each component. Finally, among the reconstructed components in each frequency band, the component with the lowest instantaneous frequency energy fluctuation is selected as the optimal component of each frequency band, and is taken as the final adaptive generalized intrinsic mode function (AGIMF) of AGEWT. The proposed AGEWT method is applied to rolling bearing signals, and the results prove its effectiveness in extracting fault features.