Degradation State Analysis of Permanent Magnet Synchronous Motor Winding Based on Multiscale Multidimensional Correlated Sample Entropy

Abstract

In this article, a degradation model for the permanent magnet synchronous machine (PMSM) windings is carried out based on multiscale multidimensional correlated sample entropy (MMCSE). First, the calculation process of multiscale sample entropy (mse) is optimized to extend its applicability from 1-D data to multidimensional data. Next, the range of analysis scales is determined, and the degradation model of the PMSM winding is established based on entropy values under degradation and healthy states. Then, the simulation models of the PMSM under different degradation states are constructed, and the important parameters of the algorithm are determined using the Jellyfish optimization algorithm. Finally, the experiments are presented to validate the accuracy of simulation analysis. The research results indicate that the degradation model based on MMCSE can accurately determine whether the PMSM windings have experienced performance degradation. The calculation results of the degradation model are related to machine reliability, providing a new quantitative evaluation metric for assessing changes in the PMSM reliability.