Fault Diagnosis of Rolling Bearing Based on Wavelet Packet and Fourier Analysis

Abstract

Based on primary theory of wavelet packet transformation, the signal of rolling bearing fault was processed, The signal characteristic can be effective withdrawn from the actual signal after wavelet packet decomposition and restructuring, Then, using Fourier transform, it presents an approach to get the frequency of fault signal. The experimental result proves that, the wavelet packet analysis can adaptively choose the related frequency band according to the analyzed signal characteristic. The practice example shows that this method is effective. The research results provide the theoretical foundation for rotary machine fault diagnosis, and have important practical value.