| Aging,stroke,and accidental injury have led to many patients with ankle disorders.The ankle rehabilitation robot can effectively reduce the workload of nursing staff,shorten the rehabilitation time of patients,and improve the quality of rehabilitation.In order to develop an ankle rehabilitation robot with intelligent detection and intelligent control,the following work was carried out in this paper.(1)A tandem 3-DOF ankle rehabilitation robot was designed based on the physiological characteristics of the ankle joint and industry standards for rehabilitation medical equipment.In the no-load characteristic analysis,the no-load driving moment equations in the direction of three degrees of freedom were established,and the no-load dynamics simulations were carried out in ADAMS for single-axis motion and three-axis linkage.The results show that the trend of the single-axis motion and the three-axis linkage no-load drive torque is the same,and there is only a small difference in the magnitude of the values.(2)The generation principle and acquisition scheme of surface EMG signals,joint angle signals and force moment signals were analyzed.The signal acquisition circuit was designed with STM32F103RCT6 as the main chip.The lower computer signal acquisition program was written in C language,and the upper computer signal acquisition software was designed in C++ language.The data acquisition system was finally built,and after testing,the system can realize data acquisition,display and storage.(3)Based on the Hill muscle model,an ankle joint moment calculation model was established,and the experimental protocol was designed according to the optimization requirements of the model parameter set.In the experiment,data were collected from nine subjects during dorsiflexion/toe flexion,internal rotation,and internal/external rotation movements.The data of 6 subjects were used to optimize the parameter set of the model,and the optimization method was genetic algorithm,and the accuracy of the optimized model was verified with the data of 3 subjects.The root mean square errors between the calculated and measured moments of the optimized model are all less than 1.4,therefore,the optimized model predicts the ankle joint moments better.(4)Four rehabilitation modes were designed based on ankle rehabilitation needs: passive,active,assisted and supple.For the passive mode,a PID controller was designed for closed-loop position control,and the tracking error of the trajectory is less than 3.8%.For the active mode,a PI controller was designed for current closed-loop control,and the control system input eliminated the effect of no-load drive torque,and the simulation shows that there is a certain error at the commutation.For the power-assist control,a PI controller was designed for current closed-loop control,and the power-assist target was the difference between the required torque of the ankle joint motion and the result of the ankle joint torque calculation model,with a power-assist error of approximately0.63%.For soft control,a double closed-loop controller with inner-loop current and outer-loop position was designed which can effectively avoid the situation of “hard contact”.In order to verify the reliability of the designed controller,the torque response characteristics of the impedance mode were experimentally verified with small overall error,which enables effective impedance training.The ankle rehabilitation robot system established in this paper can collect patient data and realize multiple control modes,which can effectively improve the ankle rehabilitation effect.The ankle joint moment calculation model established in this paper can be used for the control of the ankle rehabilitation robot and the evaluation of the rehabilitation status.The research achievements are of reference value for rehabilitation equipment development. |
PDF Full Text Request