Human-Robot Coordination Control of Robotic Exoskeletons by Skill Transfers

Neuromotor control experiments have shown that humans are able to adapt limb impedance to stably and properly interact with various environmental forces with dexterous skills. Inspired by these observations, in this paper, we develop a control approach in which the robot automatic control is combined with impedance control using stiffness transferred from human operator. Under the assumption of linear mapping between muscle surface electromyography signal amplitude and human arm stiffness, we employ the incremental stiffness extraction method in operational space with an improved performance by compensation of the nonlinear residual error in the mapping. The teleoperated robotic exoskeleton is able to replicate the impedance of human operator's arm, and simultaneously, compensate for external disturbances by the technique of disturbance observer. Experimental studies have been carried out to test the antidisturbance ability of the proposed approach for the pose maintenance task in the presence of both stable and unstable interactive forces. The effectiveness of the proposed approach has been demonstrated by the experimental results.