基于经验小波变换的滚动轴承故障诊断研究

s a new adaptive signal decomposition method, Empirical wavelet transform (EWT) can adaptively built band-pass filters in the frequency domain so as to construct the orthogonal wavelet functions and extract the AM-FM components that have a compact support in Fourier spectrum. Hereby, EWT is introduced in the fault diagnosis of rolling element bearing. The low frequency components of bearing fault signal is subjected to the effect of modulating by the multiple bearing resonance frequencies, the bearing resonance bands carry rich fault information. A frequency band division method of empirical wavelet transform is put forward, which could determine the frequency band boundaries reasonably and preserve the resonance band.