Short-Fault Diagnosis Method for CSI-Fed PMSM Drives Based on Dual-Scale Average Current Errors

This paper proposes a robust short-circuit fault diagnosis method based on dual-scale average current errors for the permanent magnet synchronous motor (PMSM) drive fed by the current source inverter (CSI). The method only employs the already exist motor currents and does not need additional sensors and circuits, which is simple and easy to implement. The current errors between the average values of the reference currents and that of the feedback currents in one cycle are calculated, and the ratios of the errors to the average absolute values of the reference currents are used as the diagnostic variables. In addition, two amplitude thresholds are set to determine the fault type by comparing the diagnostic variables with the amplitude thresholds. Then, the faulty switch is located according to the signs of the diagnostic variables. In order to avoid misdiagnosis in the transient processes, a time threshold as the second scale is set to enhance the robustness of the diagnostic algorithm. Finally, an experimental platform is built to verify the effectiveness and the robustness of the proposed diagnosis method.