Research On Mechanism Design And Motion Intention Recognition Of Finger Rehabilitation Robot

China has officially entered the aging society in 2022.Elderly individuals are a high-risk group for stroke incidence.More people are experiencing hand dysfunction due to the damage of nervous and cardiovascular system,which brings great burden to families and society.Moreover,the current situation in China’s rehabilitation medical industry is characterized by a high number of patients and a shortage of medical professionals.In view of the problems above,this paper develope an exoskeleton finger rehabilitation robot to meet the rehabilitation needs of patients.Firstly,based on the principles of kinematic anatomy,the biological characteristics of the hand are analyzed to determine the range of motion of the finger joints during rehabilitation,and the kinematic equations of the finger are obtained using the D-H method.A finger rehabilitation robot has been designed with mechanisms for flexion/extension and adduction/abduction of the metacarpophalangeal joints for the index and little fingers,mechanisms for flexion/extension of the metacarpophalangeal joint for the middle and ring fingers,as well as mechanisms for flexion/extension of the proximal interphalangeal joints and the thumb.Taking the index finger mechanism as an example,kinematics analysis and workspace solution are carried out.Secondly,based on the therapy of the healthy side driving the affected side in hand rehabilitation,a method for recognizing hand motion intentions based on sEMG signals is proposed.After preprocessing operations such as raw data collection,filtering,and label production,the clustering distinguishability evaluation index is used to optimize signal features,and the optimized feature matrix is input into the least squares support vector machine for classification and recognition to achieve prediction of hand motion intentions.Thirdly,aiming at the problem of poor classification ability of least squares support vector machines,an improved arithmetic optimization algorithm is proposed based on integrating multiple strategies such as chaotic map initialization,sinusoidal function dynamic boundaries,and crisscross optimization algorithm,which is used for hyperparametric optimization.The proposed algorithm is tested using standard test functions to verify its convergence accuracy,speed,and robustness.Finally,the the finger rehabilitation robot system platform is built,including the hardware system,software system,and human-computer interaction interface.The joint angle testing experiments,grip performance testing experiments,and motion intention recognition ability experiments are conducted to verify the reliability and safety of the robot.