Human-Cooperative Control of a Wearable Walking Exoskeleton for Enhancing Climbing Stair Activities

Abstract

In this paper, a human-cooperative control strategy is proposed for locomotion assistance along a physiologically meaningful path through a wearable walking exoskeleton. First, a kinematics model for climbing stairs with the step length, walking speed, and direction is developed. Then, the trajectory of the center of gravity (CoG) in both sagittal plane (SP) and coronal plane (CP) of the human–robot system is designed using virtual slope method (VSM) and dual length linear inverted pendulum model (DLLIPM), which are generated in parallel to the gradient vector of the virtual slope and the horizontal plane, respectively. Then, the task workspace of climbing stairs can be divided into a human domination region and a robot assistance region through a novel designed barrier energy function, such that the motion of human's legs can be constrained within a compliant region around the desired trajectories. Moreover, based on a smooth transition between the human and robot regions, an adaptive controller is designed to simultaneously incorporate robot motion and human's capabilities. Actual experiments involving several human subjects have been conducted and its results demonstrate the performance of the control strategy.