Fault Diagnosis and Fault-Tolerant Control of a Joystick-Controlled Wheelchair

This paper presents a method of fault diagnosis and fault-tolerant control for a nonholonomic powered wheelchair. Hard faults of sensors and actuators in two drive/steering units of the wheelchair are handled. The fault diagnosis is based on the interacting multiple-model (IMM) estimator. In order to improve fault decisions, we implement mode probability averaging and heuristic decision-making rule in the IMM-based algorithm. A fault-tolerant controller designed based on Ackerman geometry enables safe motion of the wheelchair even if sensors and actuators have partially failed. Experimental results verify the proposed method.