A model-based fault diagnosis of powered wheelchair

This paper describes a method of fault diagnosis of internal sensors and actuators for a powered wheelchair. We handle hard fault and scale fault of three sensors (two wheel- resolvers and one gyro) as well as hard fault of two wheel- motors. The hard fault of the gyro is diagnosed based on mode probability estimated with interacting multi-model estimator. The hard fault of the wheel-resolvers and motors is diagnosed based on the gyro output as well as the mode probability. The scale fault of the sensors is diagnosed based on estimating the wheelchair velocity with in-vehicle laser range sensor via a scan matching method. The iterative closest point algorithm is applied for the robust estimate of the wheelchair velocity in usual indoor environments. Experimental results validate the effectiveness of our method.