Gain-Scheduled Fault Diagnosis of In-Wheel Motor Electric Vehicles

This paper presents a gain-scheduled fault diagnosis approach for in-wheel motor (IWM) driven electric vehicles. The dynamics of vehicle slip ratio, which is usually used for vehicle stability and torque allocation, is modeled to monitor the motor torque and evaluate the loss of effectiveness. A gain-scheduled extended state observer (GS-ESO) is investigated to estimate the fault in each of in-wheel motors. With the aid of gain-scheduled techniques, the formulation of fault diagnosis for IWM electric vehicles is transformed to an output feedback gain-scheduled robust control problem. Simulation results using a high-fidelity full-vehicle model in VeDYNA validate the effectiveness of the proposed approach.