A Model-Based Strategy With Robust Parameter Mismatch for Online HRC Diagnosis and Location in PMSM Drive System

Abstract

This article proposes a model-based strategy with robust parameter mismatch for online high resistance connection (HRC) diagnosis and localization in surface-mounted permanent magnet synchronous machine (PMSM) drive system. The mathematical mode of the PMSM with the HRC is established in the abc stationary frame and dq rotating frame. Based on the model, the voltage distortion estimator is applied to estimate the voltage distortions caused by the HRC. According to the voltage distortions, a new fault characteristic signal is defined and is analyzed under different fault conditions. Based on the defined signal, the fault indicators being insensitive to the parameter mismatch are defined to diagnose the fault and locate the faulty phases. By the proposed method, not only the HRC can be detected, but also the fault severity can be assessed and the faulty phases can be located, where no additional equipment or measurement devices are needed. Both the simulation and experimental results verify that the proposed method can effectively achieve online HRC diagnosis and location in the PMSM drive system.