Joint stiffness tuning of exoskeleton robot H2 by tacit learning

Abstract

Joint stiffness of the exoskeleton robot is one of the most important factors to support bipedal walking. In this paper, we discuss the robot joint stiffness tuning algorithm using the bio-mimetic learning method called tacit learning. We experimentally showed that the pro- posed controller can tune the joint stiffness of the exoskeleton robot by tuning the integral gain in the controller. The walking experiment wear- ing the exoskeleton robot suggest that the stiffness tuning is applicable to control the walking speed.