Study On A SEMG-based Upper Limb Rehab-Ilitation Training System

The loss of limb function caused by stroke diseases brings a heavy burden to the family and society of patients,with the aging of the aging.Therefore,rehabilitation training after treatment of hemiplegia patients is an important determinant of patients returning to the family.The traditional way of rehabilitation is that doctors directly perform freehand training on patients,but the process labor intensity is low and low.Therefore,the research of technology combining robot technology and rehabilitation physiotherapy has important significance for the rehabilitation of patients with limb dysfunction.In this paper,a flip-type upper limb rehabilitation robot is designed for the upper limb rehabilitation training problem.The rehabilitation training can be carried out in the inclined space,which increases the training dimension.By establishing a simplified model of the upper limbs of the human body,the kinematics simulation based on the end training trajectory is used to verify the effectiveness of the rehabilitation equipment.At the same time,various training modes are provided,and the auxiliary mode based on the sEMG signal is deeply studied and finally completed.The construction of the rehabilitation system,the main contents of this topic are as follows:Firstly,the research background and significance of this topic are expounded.The research status of upper limb rehabilitation robots at home and abroad is reviewed and summarized.At the same time,the application and development of surface electromyography signals in the field of rehabilitation robots are analyzed.The prediction classification algorithm is summarized.Secondly,it combines clinical medicine and anatomical medicine to analyze the structure and movement relationship between the joints,bones and muscles of the upper limbs of the human body.At the same time,based on the shortcomings of the current desktop upper limb rehabilitation robot,a design scheme of the inverted mechanical structure is proposed and complete the modeling of the whole machine combine design requirements,calculate and select the key transmission mechanism,and finally solve the position of the turning mechanism part,laying the foundation for the subsequent motion control.Then further verify the rationality of the design of the flipping mechanism,establish a simplified model for the upper limbs of the human body,and perform forward and inverse kinematics analysis on the simplified theoretical model.The simulation model is built by the Robotics Toolbox under the Matlab platform,and then based on different The change of the end trajectory in the angled slope is used to solve the movement of the shoulder and elbow joint angle,and the simulation results are compared and verified to illustrate the feasibility and effectiveness of the training of the flip-type rehabilitation training equipment.Finally,for the realization of the auxiliary mode,the joint angle prediction research based on the EMG signal is carried out,and the mapping relationship model between the joint motion angle and the extracted eigenvalue is established.The EMG signal characteristic value is taken as the input,and the joint angle is taken as the output.The nonlinear auto-regressive neural network with Exogenous Inputs(Nonlinear Auto-Regressive Neural Network with Exogenous Inputs,NARX)algorithm is used to construct the prediction models of the two,and compare them with the back propagation(BP)and radial basis function(RBF)network algorithms.It turns out that the prediction accuracy of NARX is the best and the stability is better.The hardware structure of the upper limb rehabilitation robot system was built,the upper computer interface and the lower computer control program were programmed,and the basic functions were realized by using the rehabilitation robot experimental platform to verify the feasibility of the rehabilitation system.