Research On Control Method Of Active Upper Limb Rehabilitation Training Robot

According to the relevant mathematical statistics,the elderly over the age of 60 have accounted for 17.3% of the total population in China.In accordance with international standards,China has officially entered the aging society.The mortality rate of cardiovascular disease,stroke and cancer is extremely high in the elderly social group,while cerebral apoplexy(stroke)caused by cerebral ischemia and cerebral hemorrhage is the most serious.Patients have a great probability of having corresponding degree of unilateral limb movement disorder(hemiplegia)during and after the disease,which have a strong impact on their daily life.Traditional rehabilitation not only has shortcomings in training methods,but also can not be real-time detection of the patient’s limbs,so through rehabilitation training robot assisted training patients for rehabilitation exercise has become the general trend.Through active and passive training tasks,the rehabilitation training robot can gradually improve the movement function of the affected limb of the patient,and help the patient establish the connection between the damaged nerve and the limb of the hemiplegia side.The research on the movement control method and system of the rehabilitation robot has broad research value and good social effect in the modern rehabilitation medicine.Aiming at stroke patients with upper limb hemiplegia,thesis designed a set of active upper limb rehabilitation training method and manipulator arm control system with 7-DOF Barrett-WAM mechanical arm.Firstly,the forward/inverse kinematics and dynamics models of the mechanical arm were modeled and analyzed.A joint Angle was fixed when the inverse was obtained,and the geometric method and analytical method were used to solve the simplified calculation process simultaneously,solve the following problems of manipulator motion control.Then based on the impedance control design active upper limb rehabilitation training robot control system,has a certain upper limb motor ability of patients by push and pull the mechanical arm end contact ball training task,system through online identification algorithm in the process of training state of real-time detection of limb,and through the resistance adaptive adjust gradually change expectations interaction force,To reach the maximum resistance that the affected limb can overcome.Finally,an experimental platform of active rehabilitation robot built by Barrett-WAM mechanical arm was used to verify the effectiveness of the proposed method.In the later stage of rehabilitation training,the static/dynamic interaction force prediction method combining s EMG signals was proposed,and the neural network prediction model was trained through the real-time information of upper limb s EMG signals and interaction forces of healthy recovering subjects.During the rehabilitation training of patients,the s EMG signals of patients are collected,and the network model after training is used to predict the interaction force of patients.By comparing with the real interaction force information,the difference of interaction force between patients and healthy subjects can be calculated,which can be used as the evaluation standard of rehabilitation training.