A Current Residual-Based Mixed Fault Diagnosis and Differentiation of PMSM Servo Systems

During the long-term operation of permanent magnet synchronous motor (PMSM) servo systems, various faults may occur, leading to system performance degradation or even endangering normal operation. Static eccentricity fault (SEF) and interturn short-circuit fault (ISCF) are two common imbalance faults that exhibit similar fault characteristics, making it difficult for general single-fault diagnosis methods to differentiate between them. Moreover, fault diagnosis methods designed for steady-state may not be suitable for servo operations. To address these issues, this paper analyzes the fault characteristics of SEF and ISCF and proposes a current residual-based mixed fault diagnosis and differentiation (CR-MFDD) method. By generating current residuals through the closed-loop observer and applying double normalization on the amplitude and frequency of its harmonic and dc components, reliable fault diagnosis and differentiation are achieved under servo conditions. Experimental validation on two motor platforms confirms the effectiveness of proposed method.