A New Control Method for Power-assisted Wheel Chair based on the Surface Myoelectric Signal

In this paper, we design a new controller of power- assist wheelchair. Conventional power-assist controllers only focus on the amplification of the human input force using torque sensor. Our proposed controller does not use a torque sensor but an electromyogram sensor and a disturbance observer. Then we decide assist torque by combining the filtered myoelectric signal and the estimated human torque signal calculated by the disturbance observer. Our designed controller carries out different power-assist method for acceleration and deceleration phase. During the acceleration phase, we provide power assist earlier than the driver´s action. On the other hand, when the driver wants to stop the wheel chair, we step down assist torque according to the degree of urgency estimated by the filtered myoelectric signal. Experimental results show that the proposed power-assist method realizes fast, stable and smooth acceleration and safe deceleration.