Research On Error Modeling And Compensation Of Cable-Driven Lower Limb Rehabilitation Robot

The cable-driven robot is a special type of parallel robot that uses flexible cables instead of rigid links.It has the advantages of simple structure,large working space,and easy reconfiguration.However,the rope is a flexible part,which causes the cable-driven robot to easily produce motion errors during the movement.Therefore,error analysis and compensation for the cable-driven robot is indispensable.The cable-driven robot in this thesis is a rehabilitation robot in which four cables are used to drive the lower limbs of the human body for gait training.It is also named a cable-driven lower limb rehabilitation robot.In order to improve the motion accuracy of the robot,the full-parameter geometric error model and error sensitivity model are establish,and then the main error sources based on the sensitivity coefficients of each error source are filtered.Finally,the main error sources are identified and compensated to improve the accuracy of the robot’s motion.First,the human gait trajectory was collected,and the configuration of the robot was optimized.The kinematics of the robot is analyzed.An optical three-dimensional measurement system is utilized to obtain the gait trajectory of a person’s normal walking.Then,the structure parameters of the robot is optimized based on the gait trajectory and maximize the working space.The third step is based on the configuration,the kinematics analysis of the robot,and the CASPR(Cable Robot Analysis and Simulation Platform for Research)software is used to verify the kinematics analysis of the optimized configuration robot.Second,optimization the gait trajectory of the joint angle.Based on the robot completing the rehabilitation training,the differential evolution algorithm is used to trace the theoretical joint trajectory,and reduce the energy.Reducing the energy consumption caused by the vibration during the movement and improving the operating efficiency and stability of the robot,the trajectory of the joint frame of the cable-driven robot is optimized.Then,a full-parameter geometric error model and sensitivity model of the robot are established.A full-parameter geometric error model of the robot is established based on the kinematics model.The sensitivity model of the robot is constructed based on the error model,and the error sources are filtered according to the global sensitivity coefficient of each error source.18 main geometric error sources are found out.Finally,the correctness of the error model and sensitivity model was verified by MATLAB(Matrix Laboratory)simulation.Finally,the main geometric error sources are divided into static error and dynamic error,and compensation is carried out.For static errors,the method of kinematics calibration is used to compensate.That is,on the basis of the error model,the actual pose of the measuring point is obtained through the optical three-dimensional motion capture system,and the least square method is used to identify the error source and correct the kinematic static geometry.ADAMS simulation result verifies the feasibility of the kinematic calibration method.The real-time compensation of dynamic errors is achieved through the optical three-dimensional measurement system and the Leisai control system.Finally,experiments verify the effectiveness of the compensation algorithm.