Torque balancing for power assisted wheelchair based on torque and temporal similarity

Abstract

In this study, a new torque balancing technique is proposed to improve the driving performance of power assisted wheelchairs (PAWs). For a PAW to maintain a straight driving direction, the same force must be delivered to both the right and the left push-rims. However, humans typically have different strength in the right and left arms, and therefore, they cannot control the force uniformly at all times. As a result, the driving direction may need frequent corrections during PAW operation. Toward this end, appropriate assist torques need to be generated to compensate for unbalanced human torque inputs by detecting the driving intention in order to maintain a uniform direction in straight sections or to quickly change direction in curved sections. We propose an assist torque balancing control scheme that estimates the human driving intention by considering the temporal similarity between the right and the left torque signals as well as the proportion of both torques. Simulation results show that the proposed balancing control scheme has better performance than the existing method for straight and circular driving conditions.