Research On Elbow Rehabilitation Exoskeleton System Based On SEMG Variable Impedance Control

The movement of the elbow joint plays an indispensable role in our daily life,and the rehabilitation training for patients with dyskinesia of the elbow joint is extremely critical.Exoskeleton rehabilitation robots have received enormous social attention and recognition in recent years owing to their good rehabilitation effect and low cost.From the point of the structural design and rehabilitation strategy,a systematic study on elbow rehabilitation exoskeleton based on myoelectric signal control was carried out.The content of the research mainly includes the structural design of the elbow joint exoskeleton robot,the modeling study on s EMG-human body impedance parameter,the active and passive rehabilitation control strategies and corresponding simulation experimentsThrough the anatomy,biomechanics and clinical case analysis of the human elbow joint,the design requirements of the elbow joint rehabilitation exoskeleton were clarified.Considering the human-machine coupling state,the structure and transmission method of the three-degree-of-freedom exoskeleton are designed.Especially to solve the problem of man-machine axis misalignment,a misalignment compensation mechanism was designed,compensating the problem of axis misalignment and axis deflection,ensuring that the exoskeleton was light and easy to wear,adjustable in size and flexible in transmission.The prototype was developed.The research on human body impedance modeling based on EMG signals was completed.The reference muscle activation model,muscle contraction model as well as simplified musculoskeletal model of upper limb was established.An experimental platform for measuring multi-source information of s EMG signals,joint angles,and joint torque was built,and the measurement experiment was designed to extract elbow joint stiffness information.The stiffness trend parameter identification of the musculoskeletal model was carried out through stiffness information,and the model was corrected through repeated experiments with multiple joint angles.A joint stiffness model based on EMG signals was established.Aiming at the different stages of rehabilitation,an advanced control strategy for elbow rehabilitation robots was studied.Based on the fuzzy PID algorithm,a fuzzy PID controller was designed to follow the patient's trajectory to complete the passive rehabilitation training.A variable impedance control strategy based on s EMG was proposed,and the effect of impedance parameters on the control result was simulated and verified,realizing the training strategy of the active variable impedance rehabilitation stage.Using impedance control to achieve force tracking,the training strategy in the resistance rehabilitation stage was designed.A model for simulation was built based on Adams and Matlab.The simulation verifies the effect of the passive compensation mechanism.Meanwhile,passive rehabilitation strategy,active rehabilitation strategy and resistance rehabilitation strategy was verified through the individual experiment,where the rationality of structural design,the feasibility of control strategy and the s EMG-based variable impedance control were proved to be effective.