Robust Online Diagnosis of Inverter Open-circuit Switching Faults for Robotic Joints with BLDC Motors

Online health monitoring of robotic joint motor drive systems is of paramount importance for robots utilized in remote, safety-critical, and hazardous environments. Joint motors and the associated power electronic drives are prone to hardware failures in harsh environments and may yield joint failures. Brushless DC (BLDC) motors and the related inverters have been considerably used in robotic arms; therefore, numerous inverter fault diagnostic strategies for BLDC motor drives have been proposed in recent literature to enhance the system's reliability and avoid downtime cost. In this paper, a robust fault diagnosis technique of semiconductor switches is proposed for BLDC motor drive systems in robotic applications based on the stator current signature analysis. The performance of the BLDC joint motors is investigated under inverter switches' open-circuit faults using finite element (FE) co-simulation tools. Besides the robust fault diagnosis capability, the proposed methodology is capable of identifying the faulty switch based on a knowledge table by considering several fault conditions. The robustness of the proposed technique has been verified through extensive simulations under numerous speed and load scenarios. Finally, theoretical findings are validated using the Kinova Gen3 robot arm to highlight the effect of joint faults on the robot's reachability of desired task locations.