Research On Adaptive Stability Control Method Of Upper Limb Exoskeleton Rehabilitation Robot

In recent years,with the continuous increase of the aging population,the number of limb motor dysfunction caused by stroke,accidental injury and other reasons is also on the rise,which poses a huge challenge to the allocation of medical resources in the whole society.The most effective treatment for hemiplegic patients today is long-term rehabilitation,which treats the damaged muscles and peripheral nerves to restore some or most of the original motor function.With the gradual development of rehabilitation robots,which not only provide precise and quantifiable training,but also overcome the shortage of medical staff,so robot-assisted rehabilitation of the affected limb has become a trend in the future of healthcare.The rehabilitation robot is required to follow a predetermined trajectory when tugging the patient’s limb for recovery exercises,but it is often faced with various external disturbances that make it difficult to follow the desired trajectory accurately.In addition,as the strength of the patient’s limb increases,uncontrollable muscle activity is generated,which affects the stability of human-robot interaction and is detrimental to the patient’s rehabilitation training.In this paper,the problems of upper limb rehabilitation robots are discussed in depth,and the main research work and content are as follows:(1)The main research direction of this paper is to analyse the development of rehabilitation robots at home and abroad in recent years,to discuss the types and applications of control strategies during the various training phases of rehabilitation robots,and to propose the main research directions of this paper in the light of the current research situation.(2)Combined with human bionics,a master-slave upper limb exoskeleton rehabilitation robot is designed.To ensure the control and execution of rehabilitation training,the kinematic model of the robot is established by improving the D-H parametric modelling method,the kinetic equations of the robot are derived using the Lagrangian method,the master-slave mapping model is designed,and the Simscape model of the robot is validated to complete the visual simulation of chest expansion training.(3)For the passive rehabilitation phase of the patient,a PID control algorithm is designed to control the joint trajectory directly based on the kinematic model of the robot,but this method has poor immunity to disturbance,so a fuzzy adaptive PID control algorithm is designed to improve it.In addition,based on the robot dynamics model,an adaptive control method based on RBF neural network with sliding mode robust terms is designed to improve the adaptivity of the system,as the common computational torque control strategy cannot overcome the modelling error and joint friction interference,and the two adaptive control methods are validated by simulation respectively.(4)For the patient interaction training phase,a position-based impedance control model is designed to analyze the influence of impedance parameters on the response of the robot system by studying the human-machine contact impedance.And in order to ensure the stability of human-machine interaction,the design model refers to the adaptive impedance controller to improve the robustness of the system,and combined with simulation to prove that in the presence of external disturbances,the proposed method can achieve good trajectory tracking and contact force tracking compared to position impedance control,adapt to external changes,and achieve high control accuracy.This study focuses on the control strategy of an upper limb exoskeleton rehabilitation robot.When a patient wears a rehabilitation robot for training,it is susceptible to various types of interference such as joint friction,modelling errors and changes in human-machine contact forces,resulting in reduced trajectory following accuracy and thus affecting the patient’s rehabilitation training.In order to improve the anti-interference performance of the rehabilitation robot,the kinematic and kinetic models of the robot are used as a basis to design an adaptive stabilisation control strategy for each stage of the patient’s recovery training.