Minimum jerk control of power assisting robot on human arm behavior characteristic

Abstract

This paper presents a novel trajectory control-method for power assisting robot. Human input force is intermittent in many cases such as assisting devices for upper limbs and power assisting wheelchair; therefore, the suitable trajectory must be generated also after the human decreases his force. This paper tries to solve this significant problem based on minimum jerk model closely approximating human arm movements because smooth and human-friendly power assisting robot can be realized by imitating human arm behavior characteristic. Some experiments and evaluation by trial subjects on power assisting robot show the effectiveness of the proposed method