Application of wavelet transform in fault diagnosis of rolling bearing

In order to detect the fault signal of rolling bearing, the fault diagnosis of rolling bearings is carried out by using discrete wavelet transform. The practical vibration speed signals measured from rolling bearings are decomposed and reconstructed by Mallat algorithm. Then an envelope analysis is made to the signal. The fault of rolling bearing component is diagnosed by extracting fault feature from envelop frequency spectrum figure. The experiments results showed that mutation signal can be easily found from detail signals after N-decomposition of the vibration signal of rolling bearing. The existence of fault points can be judged accurately by detecting the characteristic frequency of fault signals from the power spectrum after Hilbert envelop.