A novel open-circuit fault diagnosis method for BLDCM inverters based on current-related characteristics

Abstract Inverter switch failure is a critical issue in brushless DC motor (BLDCM) systems, significantly reducing production efficiency and increasing maintenance costs. This paper proposes a multi-switch fault diagnosis method for six-step commutation-controlled BLDCM inverters, leveraging periodic current-related characteristics within a sliding window. First, the types of open-circuit (OC) faults in multiple inverter switches are investigated and their mechanisms are analyzed. Second, four basic feature variables are extracted from the three-phase current signals within the sliding window, which can be used to determine the position of the fault switch through simple threshold judgment and mathematical operations. Finally, simulation and experimental results verify that the proposed diagnostic method is reliable, effective, and robust. Moreover, this method can detect 21 types of OC switch faults without additional sensors and can be embedded in microcontrollers for practical engineering applications.