Uncertain Pushing Adaptive Coordinated Control for the Human-Exoskeleton-Walker System

Abstract

Lower Limb Exoskeletons are potential in the gait training for patients with gait disorders. For patients in the early rehabilitation stages with weak upper limb strength, it is challenge to keep balance by themselves only. A mobile robotic walker is helpful to maintain the walking balance, with the help of a physician pushing behind to keep a forward walking associated with the exoskeleton. However, since the gait patterns are varying with different training tasks, how to ensure a coordinated movement between the exoskeleton and the mobile robotic walker is challenge, especially with the uncertain pushing applied by the physician. In this letter, the Uncertain Pushing Adaptive Coordinated Control (UP-ACC) approach is proposed to solve the problem, which consists of two contributions. Based on the decoupled simplified dynamics, the center of mass trajectories and footstep placements are generated through linear model predictive control. In addition, the optimal footstep placement is used for the human-like gait planning by introducing the phases of heel-strike and toe-off. The proposed method has been verified in the robot simulation platform CoppeliaSim, and the experimental results indicate its effectiveness in generating coordinated motion and human-like gait patterns for the human-exoskeleton-walker system with the external pushing from 0 to 150 N and walking speed from 0.2 m/s to 0.6 m/s.