Human-Robot Interaction Force based Power Assistive Algorithm of Upper Limb Exoskeleton Robots Driven by a Series Elastic Actuator

Abstract

Upper limb exoskeleton robots have been widely used to assist humans in industry and rehabilitation. Human-robot interaction control consists of two modes. A compliant mode and an assistive mode. The compliant mode is manipulating the robot compliantly according to the movement of the human. At this time, the human must not receive impedance from the robot. Therefore, in this paper, a novel external force control which makes the robot follow human movement without causing resistance is proposed. The assistive mode is to generate assistive power according to human intention. In this paper, the algorithm for recognizing the human intention for primitive movement and generating assistive forces adaptive to real-time work environments is proposed. The proposed two modes are also integrated into a novel human-robot interaction control framework. The performance of the proposed methods is verified through experiments.