Robotic Knee Tracking Control to Mimic the Intact Human Knee Profile Based on Actor-Critic Reinforcement Learning

作     者：Ruofan Wu Zhikai Yao Jennie Si He(Helen)Huang 

作者机构：School of ElectricalComputer and Energy EngineeringArizona State UniversityTempeAZ 85287 USA Department of Biomedical EngineeringNorth Carolina State UniversityRaleighNC 27695 USA University of North Carolina at Chapel HillChapel HillNC 27599 USA 

出 版 物：《IEEE/CAA Journal of Automatica Sinica》 (自动化学报（英文版）)

年 卷 期：2022年第9卷第1期

页      面：19-30页

核心收录：

学科分类：080202[工学-机械电子工程] 08[工学] 0804[工学-仪器科学与技术] 0802[工学-机械工程] 

基　　金：This work was partly supported by the National Science Foundation(1563921 1808752 1563454 1808898) 

主　　题：Automatic tracking of intact knee configuration of robotic knee prosthesis direct heuristic dynamic programming(dHDP) reinforcement learning control 

摘      要：We address a state-of-the-art reinforcement learning(RL)control approach to automatically configure robotic pros-thesis impedance parameters to enable end-to-end,continuous locomotion intended for transfemoral amputee ***,our actor-critic based RL provides tracking control of a robotic knee prosthesis to mimic the intact knee *** is a significant advance from our previous RL based automatic tuning of prosthesis control parameters which have centered on regulation control with a designer prescribed robotic knee profile as the *** addition to presenting the tracking control algorithm based on direct heuristic dynamic programming(dHDP),we provide a control performance guarantee including the case of constrained *** show that our proposed tracking control possesses several important properties,such as weight convergence of the learning networks,Bellman(sub)optimality of the cost-to-go value function and control input,and practical stability of the human-robot *** further provide a systematic simulation of the proposed tracking control using a realistic human-robot system simulator,the OpenSim,to emulate how the dHDP enables level ground walking,walking on different terrains and at different *** results show that our proposed dHDP based tracking control is not only theoretically suitable,but also practically useful.