| In the rehabilitation of stroke patients,the problem of upper limb movement disorders is more and more common,and it is difficult to deal with the heavy physical rehabilitation training task only relying on rehabilitation medical staff.The application of robot technology in upper limb rehabilitation training of stroke patients can greatly reduce the work intensity of rehabilitation medical staff,and can effectively meet the increasing needs of rehabilitation training.In this paper,the optimal rehabilitation position of the patient’s upper limb relative to the cooperative robot,the prediction of human-robot interaction force and the compliant control strategy of the human-robot system are studied in the collaborative robot traction method for upper limb rehabilitation.The main research contents are as follows:(1)Parameter identification of man-robot system and recommendation of optimal rehabilitation location.The characteristics of rehabilitation movement of human-robot system were analyzed,and the upper limb kinematics model was simplified to meet the basic needs of rehabilitation movement of human upper limb,and a simplified upper limb kinematics model with four degrees of freedom was established.Through the standardized design of the human-robot position,and based on the traction motion of the cooperative robot along the set trajectory,the identification method of the kinematics parameters of the upper limb and humanmachine position parameters was proposed,and the effectiveness of the method was verified by experiments based on the actual human-robot cooperation system.Based on the analysis of the spatial characteristics of human upper limb movement and the analysis of the dexterity of cooperative robots,the index of human-robot position matching degree was proposed,and the optimal rehabilitation position of human upper limb relative to the cooperative robot was recommended.(2)Identification of human upper limb dynamics parameters based on traction feedback.The Lagrange method was used to establish a simplified upper limb dynamic model,and the equivalent conversion method of upper limb traction drive and joint torque drive was proposed.The identification model of human upper limb dynamics parameters based on traction feedback was established,and an improved particle swarm optimization algorithm was proposed to realize the identification of upper limb dynamics parameters.The effectiveness of the method was verified by simulation and practical human-robot collaborative motion experiments.At the same time,the recognition effect was compared with the traditional particle swarm optimization algorithm and recursive least squares algorithm.(3)Study on compliance control method of upper limb rehabilitation man-robot system based on cooperative robot.Based on the impedance control method,a compliant control model of man-robot collaboration was established,and the human-robot interaction force in the impedance model was predicted by the simplified upper limb dynamics model.The influence of impedance parameters on the established compliance control system was analyzed.The traditional impedance control method was improved by combining with BP neural network method,the selection of impedance parameters was optimized,and the effectiveness of the method was verified. |
