Online Diagnosis of Short-Circuit Faults of Permanent Magnet Synchronous Generator by Short-Time Analysis of the Three Phase Amplitude-Phase Signal Based on Analytical Modeling

Turn-to-turn short-circuit (TTSC) fault is one of the most common faults in electric machines and may destroy coil insulation and demagnetize the magnets. In addition to the TTSC fault, this paper also introduces and analyzes the phase-to-phase short-circuit (PPSC) fault, which can have even more destructive effects than the TTSC fault. The analytical equivalent magnetic network (EMN) method is used to accurately model the healthy and faulty machine with a shorter processing time than the finite element method (FEM). The output current of the generator in the $\alpha \beta $ frame is employed to detect the faults. The mathematical equations demonstrate the change in amplitude and phase of the signals under fault conditions. Therefore, the short-time analysis of the amplitude-phase signal (ST-APS) method is presented, and the fault index is obtained by extracting real-time amplitude and phase. In addition, the proposed method separates TTSC and PPSC faults. Finally, FEM and EMN modeling results are compared with the experimental results for the prototype permanent magnet synchronous machine (PMSM). The results showed that the proposed EMN method combined with ST-APS achieves a high accuracy and enables fault detection in various stages, regardless of changes in generator speed or transient conditions. Furthermore, the proposed fault detection method is unaffected by various nonlinear loads or winding configurations, ensuring high reliability.