Bionic Hand Motion Control Method Based on Imitation of Human Hand Movements and Reinforcement Learning  

作　　者：Jibo Bai Baojiang Li Xichao Wang Haiyan Wang Yuting Guo 

机构地区：[1]The School of Electrical Engineering,Shanghai DianJi University,Shanghai,201306,China

出　　处：《Journal of Bionic Engineering》2024年第2期764-777,共14页仿生工程学报（英文版）

基　　金：supported by Su Yan Yuan(“Development and industrialization of intelligent multi-degree-of-freedom arm based on perceptual fusion and collaborative control"(Su Yan Yuan[2019]No.107));Shanghai DianJi University(“Research on flexible joint and adaptive control technology for new upper limb prosthesis"(scientific research start-up fund project of Shanghai DianJi University)and“Research on robot intelligent grasping technology based on visual touch fusion in unstructured environment"(Science and technology[2020]No.79 of Shanghai DianJi University)).

摘　　要：Bionic hands are promising devices for assisting individuals with hand disabilities in rehabilitation robotics.Controlled primarily by bioelectrical signals such as myoelectricity and EEG,these hands can compensate for lost hand functions.However,developing model-based controllers for bionic hands is challenging and time-consuming due to varying control parameters and unknown application environments.To address these challenges,we propose a model-free approach using reinforcement learning(RL)for designing bionic hand controllers.Our method involves mimicking real human hand motion with the bionic hand and employing a human hand motion decomposition technique to learn complex motions from simpler ones.This approach significantly reduces the training time required.By utilizing real human hand motion data,we design a multidimensional sampling proximal policy optimization(PPO)algorithm that enables efficient motion control of the bionic hand.To validate the effectiveness of our approach,we compare it against advanced baseline methods.The results demonstrate the quick learning capabilities and high control success rate of our method.

关 键 词：Bionic hand Reinforcement learning Motion decomposition Multidimensional sampling PPO algorithm 

分 类 号：Q811[生物学—生物工程]