Adaptive gait generation based on pose graph optimization for Lower-limb Rehabilitation Exoskeleton Robot

Abstract

The lower extremity rehabilitation exoskeleton robot can achieve rich functions through well-designed software and hardware control systems, and bring the gospel to patients with stroke in society. Exoskeleton in the current market has problems such as poor human-computer interaction and a single rehabilitation training scene. Based on the exoskeleton system, this subject studies the human, machine, and environment interactive control strategies of the independent exoskeleton and the adaptive weight-reduction system of the exo-skeleton rehabilitation robot. These studies will significantly improve the human-computer interaction of exoskeleton and expand the rehabilitation training scene of exoskeleton. The innovation of this subject is to design a complete set of independent exoskeleton software and hardware systems, and to design an efficient real-time sensing algorithm for the independent exoskeleton system, and to propose an adaptive trajectory generation algorithm that can adapt to different terrain walking to complete Human, machine, and environment interactions of the exoskeleton system. For the bench-type exoskeleton system, we designed an adaptive weight loss system based on force control to help patients with severe illness to complete rehabilitation training.