284 / 2021-04-16 22:00:00

A bearing fault detection methodology based on enhanced adaptive resonance technique

Resonant frequency, Power spectrum, optimal band-pass filter, Rolling bearing, Composite fault diagnosis.

Machine condition monitoring and fault diagnosis

According to the following situations: (1) The being shortcomings of the existing signal processing methods; (2) The resonance frequency is obtained on the basis of the Fourier spectrum in the traditional high-frequency resonance technique (HFRT), and the bandwidth is based on experience, which is often submerged by noise and the effect is unsatisfactory. A new bearing fault detection methodology named enhanced adaptive resonance technique based on power spectrum analysis (PS-EART) is introduced to realize the single or compound fault diagnosis of the bearing in power devices. The idea is described as: first, the resonance frequency is obtained by performing power spectrum analysis on the original signal, which is used as the center frequency. Then, the bandwidth formula of the band pass filter is given, and the bandwidth coefficient is optimized using the envelope kurtosis index. Finally, the original signal is denoised by the optimized band-pass filter, and then the envelope power spectrum analysis is used to the filtered signal to extract the bearing fault characteristic frequency to realize fault diagnosis.