Signal-Based Position Sensor Fault Diagnosis Applied to PMSM Drives for Fault-Tolerant Operation in Electric Vehicles

Abstract

This paper presents a novel scheme for fast fault detection and isolation (FDI) of position sensors based on signal processing and fault-tolerant control (FTC) for speed tracking of an electric vehicle (EV) propelled by a permanent magnet synchronous motor (PMSM). The fault is detected using a comparison algorithm between the measured and delayed rotor speed signals. The proposed scheme is more practical for diagnosing faults over a wide speed range since it does not use estimated speed value. In addition, to ensure continuous vehicle propulsion and to retain effective field-oriented control of the EV-PMSM in the event of a fault, a reconfiguration mechanism with back-EMF based position observer is employed. Rapid detection of position sensor failure is necessary for a seamless transition from sensored to sensorless control. Furthermore, a comparative analysis between sliding mode observer and flux observer for motor speed control is also presented in the context of EVs. The effectiveness of the position sensors for FDI and FTC is validated in the presence of typical vehicular disturbances, such as uneven road conditions and wind disturbance force. Finally, to validate the proposed approach experimentally in a real-world EV environment, this paper utilizes a scaled-down testbed with a TMS320F28379D DSP for the motor control of the EV.