| In recent years,the cable-driven parallel robot has made a lot of research achievements in the field of rehabilitation.The cable-driven rehabilitation robot uses flexible cables to pull the patient’s limbs for movement,and has the characteristics of large working space,low inertia,good flexibility,high safety factor,so it has received wide attention from all walks of life.Based on the cable-driven rehabilitation robot independently developed in the laboratory,this subject established its kinematics and dynamics models,planned effective rehabilitation training motion trajectory and designed the corresponding controller for this system,and studied the recognition of motion intention based on the surface electromyography(s EMG)of patients’ lower limbs,so as to help patients achieve feasible and effective rehabilitation training.The main research contents are as follows:Firstly,the physiological characteristics of patients’ lower limbs were analyzed,and the forward and inverse kinematics models of patients’ lower limbs meeting the joint constraints and rehabilitation training needs were established.A direct solution method based on space geometry was proposed to realize the direct mapping from the ankle joint trajectory of patients’ lower limbs to the lower leg trajectory.Considering the reconfigurability of the system,the kinematics model of the cable-driven rehabilitation robot was established.According to the reasonable cable distribution method,the lower leg of the patient was regarded as the mobile platform of the cabledriven rehabilitation robot,and the mapping between the end point trajectory of the lower limb of the patient and the cable length change curve was realized.Taking flexion knee joint rehabilitation action as an example,combined with Monte Carlo method,adaptive s-curve and other methods,the trajectory of its action was planned,which provides the desired tracking curve for controller design.Secondly,in order to give full play to the advantages of human-machine interaction and improve the enthusiasm and initiative of patients in rehabilitation training,taking advantage of the characteristic that the s EMG of the human lower limbs is ahead of the actual action,the use of surface electromyography signals to obtain the movement intention of patients’ lower limbs was studied,so that the autonomous switching of the cable-driven rehabilitation robot system among 7different lower limb rehabilitation actions was realized.Thirdly,since the cable-driven rehabilitation robot is a typical human-machine interaction system,the dynamic model of patients’ lower limbs in the joint space and the dynamics model of the cable-driven rehabilitation robot drive system in the cable length space were established by using the Lagrange method,and a new dual-space dynamic model including the active force of the patient’s lower limb was formed.According to the characteristics of the model,considering the influence of parameter perturbation and unmodeled dynamics in the system,combined with the finite-time sliding mode disturbance observer,a desired tension feedforward impedance control algorithm and an actual tension feedback impedance control algorithm were proposed.This algorithm can not only realize passive tracking training when patients’ lower limbs are completely paralyzed,but also ensure the safety of patients during the training process when patients’ lower limbs have active power.Finally,in order to prove the rationality and effectiveness of the research content of this subject,the experimental verification was carried out.Based on the hardware system of the cable-driven rehabilitation robot,the corresponding software platform was developed.The experimental process was designed in detail,and the corresponding movement intention recognition and rehabilitation action training experiments were carried out on volunteers.The experimental results show that the research content of this subject is reasonable and effective,and can help patients achieve the expected lower limb rehabilitation training tasks. |
PDF Full Text Request