Research On Rehabilitation Training Method Of Upper Limb Rehabilitation Robot

Patients with cerebrovascular diseases and hemiplegic diseases increase year by year as the population ages,and their lives and work are severely affected.At the same time,there is a large gap in the number of doctors and patients.It is a general trend to use rehabilitation robots to assist patients in completing training.Rehabilita-tion robot can help patients to carry out stable limb rehabilitation movement,which can promote the establishment of brain damaged nerves and limb connections.The research on the control method and system of rehabilitation robot has broad scientific research value and good social effect in medical rehabilitation.In response to the arm rehabilitation needs of patients with hemiplegia in the early and late stages,this paper designs a targeted passive and active rehabilitation training method based on the 7-degree-of-freedom Barrett-WAM robot.For the dis-turbance of the control system caused by hand shake or machine friction in passive rehabilitation training,the fuzzy approximation method is used to obtain the disturb-ance amount,and the corresponding robust term is added to the system to eliminate it to ensure the flexibility of the system,and then the stability of the method is proved by Lyapunov function.Finally,the Barrett-WAM robot is used to assist the subjects to complete the passive rehabilitation experiments in the vertical plane and the horizon-tal plane,the experimental results prove the effectiveness of the method.In the later period of rehabilitation training,a fuzzy impedance control method based on variable structure compensation of patients' motion intention is proposed.The recognition of movement intention and movement was completed by collecting and processing arm EMG.In this paper,six arm movements of "elbow flexion/ex-tension","wrist buckle/abduction" and "arm internal rotation/external rotation" are designed,and the collected arm EMG signals are preprocessed and feature extraction is performed.And the IQPSO-SVM method is used to identify the arm movement intention and action,this method is to improve the optimization speed and avoid the particle optimization by adjusting the quantum particle swarm scaling factor function and introducing a certain mixed disturbance into the point of local optimal value,and finally verified by experiments that the method has good arm movement intention and action effect.In the active rehabilitation training,the difference equation of the adaptive for-getting factor is used to identify the impedance parameters of the affected limb,and then the stability of the robot arm torque and position control is improved by the im-pedance control method,and the robot system combined with the motion intention of the subjects to assist them to complete the single joint and multi joint compound ac-tive training movement.Experiments show that the trajectory of rehabilitation train-ing is relatively smooth,the interaction force between the mechanical arm and the subject is adjusted by the controller in real time.It provides assistance when the af-fected limb was not able to take initiative,and exerts resistance when there is more power,so as to achieve the effect of active rehabilitation training.