Fault Diagnosis of Motor Bearing Based on Speed Signal Kurtosis Spectrum Analysis

Abstract

Bearing failure is one of the most common types of motor faults. For the electrical diagnostic methods of permanent magnet synchronous motor(PMSM) bearing faults, the existing literatures are mostly based on the motor current signal, but subject to the limitations of sampling frequency and other factors, the current method can guarantee the analysis effect only through external signal acquisition such as oscilloscope. Considering the relatively high resolution of encoder in PMSM, proposed a diagnosis method based on motor speed with kurtosis spectrum analysis which can promote the impact component extraction to improve the signal-to-noise ratio(SNR). Compared with current method, it can promote diagnosis stability and reduce costs. In this paper, the integrated model of motor and bearing is constructed, then the simulation experiment is carried out based on the speed method. Experiments based on speed and current signals are carried out to make comparison, and the mature vibration method is used as reference. The results show that although the speed method is affected by the speed domain and load torque, the diagnostic strategy combining the rotational speed signal and the kurtosis spectrum can basically reach the effect of vibration method and has a good full-speed versatility, which is obviously better than the current signal.