Study On Control Strategy Of Lower Limb Rehabilitation Robot Based On EMG

Rehabilitation robotic system is a new technology in the field of strokerehabilitation， its appearance changes the relationship between the doctor and the patient.However, the existing rehabilitation robot control strategy is not only likely to cause thepatient’s fatigue, muscle injury and other problems, but it is difficult to mobilize theenthusiasm of the patients, this has limited the promotion of rehabilitation robot. Thisarticle is aimed at the problem that rehabilitation of existing robots is difficult to reflectthe patient’s adaptability and is lack of individual initiative. The electromyography(EMG)will be introduced into the lower limb rehabilitation robot to reflect the intent of themotion and fatigue. The control method which fuses characteristics of EMG will bediscussed. The specific works are as follows:Firstly, lower limb physical mechanism of single degree of freedom rehabilitationrobot is analyzed to get the geometric model; Geometry method and Lagrange functionmethod are used to analyse kinematics and dynamics of the man-machine system to getthe kinematics model and dynamics model.Secondly, generation mechanism of the surface EMG is analyzed; the EMG signalacquisition system is designed; based on the feature extraction method of EMG, theidentification methods of the movement intent and muscle fatigue is studied.What’s more, a rehabilitation robot control strategy based on electromyographycharacteristics is proposed: trajectory correction based on neural network is designed,robot velocity and trajectory parameters are corrected based on EMG feature of the intentand muscle fatigue, so that the awareness of the patient’s active participation and systemsecurity are raised; the robot controller based on the computed torque method is designedbased on dynamics model of man-machine system, it can achieve effective control of theman-machine system; For individual differences in patient characteristics, the PIDcontroller based on echo state network is designed, it can learn the dynamic model of theman-machine system and achieve individual control of adaptability.Finally, control experiment of the lower limb rehabilitation robot based on EMG feedback is built to study control strategy of rehabilitation by collecting EMG fromhealthy people in rehabilitation exercise. The PID controller based on echo state networkis used in tracking desired trajectory, and the effectiveness of control strategy is verifiedby the experiment.