Application Research On The Lower Limb Rehabilitation Of Multi-Channel SEMG

With the increasing levels of rehabilitation medical and the rising proportion of the disability population, the traditional manual rehabilitation has been very difficult to meet the needs of the patients, therefore the new method of combining sEMG with rehabilitation robot has gradually become a hot topic on medical research. sEMG is physiological electrical signals generated by superposition of the motor unit action potentials in time and space. The muscle generate the physiological electrical signals when human casually or involuntary motion under the domination of the nervous system. The quality of the sEMG can sometimes declare the health status of the nerves and muscles the research of sEMG characteristics can effectively identify human motion action, predict force, figure out muscle disease and direct rehabilitation. The paper focuses on the application of the lower limb rehabilitation and finish following research work:(1) Research wireless sEMG signal capture principle and design a set of high-precision, low-latency, portable wireless Wi-Fi sEMG acquisition card and figure out the schematics and software solution of sEMG acquisition card;(2) Study the methods of obtain the attitude of the human lower limb movement with sEMG, the research use AR model to extract sEMG characteristics and use SVM algorithm to classify the goble category of lower limb movement;(3) Research methods to solve the relationship between the sEMG and lower limb movement forces, the paper use SVR algorithm to figure out approximating function which described relationship between RMS of sEMG and lower limb movement force.A large number of sEMG’s RMS and actual force sample data can help estimation model to improve the estimation precision;(4) Research the relationship between movement position of6-DOF robot and sEMG in the active and passive rehabilitation training by posture prediction algorithm combined with force algorithm. In passive rehabilitation training, robot adjust position to follow the movemnet of the health side of the lower limb and help the stump to finish rehabilitation training; In active rehabilitation training, the paper introduce of the force control ring which based on the impedance control algorithm to to adjust the strength of the damping force between the robot and the limb, then complete rehabilitation robot applications based on sEMG.This paper choice the application of sEMG in the lower limb rehabilitation training as the main subject and base on the sEMG solving posture and movement strength of the movement of the lower limbs, combine with the biomechanical relations of6-DOF parallel robot platform, then complete the application research of lower limb rehabilitation.The paper figure out the method to identify the lower limb moving posture and predict the lower limb moving force by extracting of the lower limb sEMG and using Support Vector Machine, Nonlinear Autoregressive Model. At the same time, the paper work out the relationship between the motion position of robot and sEMG and use6-DOF robot combined position and impedance control method to completed the lower limb rehabilitation.