Research On Anthropomorphic Motion Planning Algorithms For Prosthetic Arm

With the increasing number of amputees all over the world,how to improve the quality of life of amputees has become a major social problem.Prosthetic limb is a device that interacts with human body.In addition to the shape and size of the prosthetic limb need to be consistent with the normal human body,it also needs to ensure the anthropomorphism of the prosthetic trajectory.In this paper,we focus on how to reproduce the natural motion trajectory of human upper limbs in prosthesis,and carry out research on anthropomorphic motion planning algorithms in combination with a seven-degree-of-freedom anthropomorphic prosthetic arm platform to provide support for reproducing the daily functional motion of human upper limbs in prosthetic arms.Therefore,the following work is done in this paper.(1)The D-H coordinate system of prosthesis arm is analyzed.By collecting the natural motion dataset of human upper limbs,based on the internal and external coordinates of the wrist position of the human upper limb motion chain,the coordinated relationship between the internal and external coordinates is established using a multilayer perceptron model.The inverse kinematics algorithm of the prosthetic arm is proposed to solve the problem that the inverse kinematics of the redundant prosthetic arm is not anthropomorphic..(2)The characteristic trajectory of human upper limb movement is extracted by probability statistical model.Combined with trajectory optimization method,a generalized humanoid motion planning algorithm is proposed The signed distance field is used for obstacle avoidance,and the weights of different components in the algorithm are optimized by covariance matrix adaptive evolution strategy(CMA-ES).(3)Based on the robot operating system(ROS),the simulation platform of the upper layer motion planning of the robot is built,and the simulation model of the prosthesis arm is established in the ROS.The motion planning algorithm proposed in this paper is integrated into the simulation platform,and finally,the effectiveness of the proposed algorithm is verified in the simulation environment.(4)The design of a variety of experimental schemes verifies that the proposed algorithm has the ability to reproducing the function of human upper limb and generalization in the new environment,and at the same time it is compared with other existing planners,which proves that the trajectory quality of the proposed algorithm is higher and more anthropomorphic.