Research On Motor Function Evaluation And Rehabilitation Control Method Based On Active Participation Level Of Joint

Rehabilitation robots can replace rehabilitators to complete a large number of repetitive movements,not only with high motion accuracy,but also achieve quantitative motor evaluation and multiple modes of rehabilitation training,which can improve rehabilitation efficiency.During the rehabilitation process of patients,it is particularly important to conduct accurate guidance training,quantitative motor function evaluation,and compliant human-robot interaction on their limb joints.Taking the rehabilitation of upper limbs as an example,although current exoskeleton robots are easy to achieve training and evaluation of upper limb joints,most of them are inconvenient to wear due to complex structures,while end-effector robots are difficult to achieve such functions due to simple structural forms and lack of modeling theory for human upper limbs.Therefore,designing an upper limb rehabilitation robot system that not only has a relatively simple structure,but also can achieve precise joint guidance training and comprehensive motor function evaluation has important application value.Firstly,a single arm upper limb rehabilitation robot system is designed and established.A method for solving the active joint torque of the upper limb is proposed.The kinematics and dynamics models of the rehabilitation robot and the upper limb of human are established using the D-H method and the Newton-Euler recursive algorithm,respectively.A perturbation observation method based on generalized momentum is used to estimate the interaction force at the end of the robot,and the transformation methods of pose and generalized force between different coordinate systems are researched,the motion information and force information of the upper limb end are calculated from the information collected by the robot.The analytical solution of upper limb inverse kinematics is derived to solve the kinematic information and active joint torques of the upper limb joints.This method can calculate the active joint torque of the upper limb through modeling analysis and information perception of the rehabilitation robot,without additional use of other sensors,and the system design is simple.Secondly,based on the current evaluation indexes of the end of the upper limb kinematics,dynamics,and joint kinematics,and combined with the solution of upper limb’s active joint torque,the Active Participation Level of joint is proposed as an evaluation index of joint dynamics to quantitatively reflect the active participation contribution of each joint.A simulation model is built in Simulink to verify the correctness of the calculation of Active Participation Level under different force conditions.An experimental platform for a single arm upper limb rehabilitation robot is established,and six subjects are recruited to conduct evaluation experiments on the ipsilateral upper limb joints of different subjects and the bilateral upper limb joints of the same subjects.The results show that human upper limb motor function can be evaluated comprehensively by using the single arm upper limb rehabilitation robot system.Finally,aiming at the joint training needs of patients,a rehabilitation training trajectory planning method for upper limb joints is proposed.Through the calculation of the forward kinematics of the upper limb and the inverse kinematics of the rehabilitation robot,the joint motion trajectory data of the robot are obtained and joint motion control is carried out.The Cartesian admittance control method is used for active compliance control of the rehabilitation robot.The stability of the discrete admittance control system is analyzed by using the Julie criterion,and the constraint relationship between the parameters in the system is determined.The Active Participation Level of patients is added to the admittance control loop,and an adaptive stiffness function is designed to enable the rehabilitation robot to dynamically adjust the level of compliance or resistance movement based on the active motion contribution of patients,so as to meet the training needs of patients with different conditions and stages.The adaptive admittance control simulation based on Active Participation Level and terminal interaction force are conducted in Simulink.Through comparison,it is proved that the control method based on Active Participation Level has better human-robot interaction performance.