Fault Detection of PMSM under Non-Stationary Conditions Based on Wavelet Transformation Combined with Distance Approach

This paper proposes a new method to detect mechanical faults of permanent magnet synchronous motors (PMSMs) under variable speed conditions. Several prior studies have proposed motor current signature analysis (MCSA) based methods for transient conditions; however, these methods have limitations because they require the characteristic frequency of the motor or they only verify the performance of the methods for a restricted time-varying region. Thus, the research outlined in this paper suggests a method for detecting motor faults using stator currents. The proposed method uses two techniques, continuous wavelet transform (CWT) and distance approach. In this method, after the influence of the non-stationary condition is reduced in the wavelet coefficients, the distance of the residual signal from the distribution of normal state is calculated. The performance of the proposed method is confirmed with the simulation result examining unbalance. From the results, the proposed method demonstrates better performance in small-load under non-stationary conditions.