Research On System Design And Control Strategy Of Upper Limb Rehabilitation Robot

With the intensification of aging in our country,the number of patients with limb motor impairment caused by geriatric diseases such as stroke has also increased dramatically.Among them,the upper limbs take on more functions in daily life,so the research on upper limb rehabilitation robots is of great significance to relieve social pressure.Based on the design and development of a six-degree-of-freedom exoskeleton upper limb rehabilitation robot,this paper conducts related design and research on the robot’s control system and control strategy.First of all,according to the requirements of sports safety and human-computer compatibility in rehabilitation training,a six-degree-of-freedom upper limb rehabilitation robot is designed to meet the rehabilitation requirements of shoulder,elbow and wrist,and the control system and human-computer interaction structure are designed and improved.Use QT to design and implement the front-end UI and back-end functions of the upper computer.Secondly,use the D-H method to model and analyze the forward kinematics and inverse kinematics of the above-mentioned manipulator.Determine the mapping relationship between the joint angle,the weight of the affected limb,and the human-computer interaction force,and use single-neuron adaptive PID to improve the position error caused by the fixed position closed-loop parameters in impedance control,propose an impedance control strategy based on joint interaction force,and prove the effectiveness of the force-position feedback control strategy for improving joint compliance through linear and circular motion simulations in the sagittal plane.Furthermore,the s EMG signal was used to directly estimate the active force,and the4-channel s EMG sensor was used to collect the s EMG data from the elbow muscles of 6volunteers.After filtering the original signal and preprocessing the feature extraction,the joint torque value obtained from the corresponding isometric contraction is input into the myoelectric-muscle strength prediction model based on LSTM(Long Short Term Memory)for training and prediction of joint muscle strength,while the data fusion simulation experiment was carried out on muscle strength and three-dimensional force.Finally,the experimental platform of the upper limb rehabilitation robot was built,and the impedance control experiments were carried out under the sagittal plane linear and circular trajectories to illustrate the effectiveness of the designed position-based impedance control in improving joint compliance.Through the estimation of dynamic muscle strength under continuous training state,it is proved that the designed algorithm has excellent generalization ability.