Design And Control Research For Lower Extremity Rehabilitation Exoskeleton Driven By Pneumatic Muscle Actuators

Motor disorders caused by strokes,spinal cord injuries,and skeletal muscle weakness severely limit the activities of daily living of the elderly and frail people.For patients with nerve injuries,robot-assisted therapies have been widely accepted because they can allow the higher intensity of rehabilitation training,save the efforts of therapists,improve the efficiency of rehabilitation training and shorten the rehabilitation processes.From the perspective of exoskeleton actuation,pneumatic muscle actuators(PMAs)are especially suitable for driving lower limb exoskeleton robots owing to their features of lightweight,high power/weight ratio,and natural compliance.Therefore,this paper develops a lower limb exoskeleton rehabilitation robot driven by PMAs.Aiming at the passive and active gait training methods,intelligent control,proxy-based trajectory tracking control,and compliance control are developed to realize the safe and compliant rehabilitation training for the PMAs-driven exoskeleton system.The main contents of this paper are summarized as follows.To address the problems that motors lack necessary compliance and the series elastic actuators require high power output,this paper designs the actuation unit of the antagonistic configuration of PMAs and the orthosis mechanism of the lower limb exoskeleton.In this way,a PMAs-driven lower limb exoskeleton is invented.By analyzing the movement law and driving principle of the orthosis,a simplified dynamical model of the PMAs-driven exoskeleton is established based on the Lagrangian function.And the effectiveness of the simplified dynamical model is verified through experimental studies.To deal with the problem of the parameter identification of the PMAs-driven exoskeleton caused by strong nonlinearities of PMAs,this paper proposes a nonlinear model predictive controller(NMPC)and a single-layer neural network(SLN)controller based on the echo state Gaussian pro-cess(ESGP)to realize the trajectory tracking control for the PMAs-driven exoskeleton.By using the Lyapunov theorem,the stability of the closed-loop system is analyzed.Also,the effectiveness of the proposed method is verified by experimental studies involved in multiple volunteers.The results show that the intelligent control technologies based on the echo state Gaussian process can realize the passive gait training of lower limb rehabilitation,and the tracking error is less than 3degrees.Given the lack of research on human-robot interaction security in trajectory tracking control,two proxy-based control methods for the PMAs-driven exoskeleton system are proposed in this paper.Firstly,to solve the problem that the parameters of the traditional proxy-based sliding mode controller(PSMC)are difficult to be properly tuned,a switch-mode firefly algorithm(SMFA)based proxy-based sliding mode control is proposed,which automatically tunes the parameters of the tra-ditional proxy-based sliding mode control.On the other hand,to solve the problem that the tradi-tional proxy-based control is difficult to prove the stability of the closed-loop system,this paper constructs the two-stage motion of the system states and the motion law of the virtual proxy on the basis of the simplified dynamical model of the PMAs-driven exoskeleton.As a result,a method called adaptive proxy-based robust control(APRC)is proposed.Integrated with the nonlinear dis-turbance observer(NDO),the disturbances/uncertainties of the system are effectively compensated,and the robustness of the system is improved.According to the Lyapunov theorem,the system states are proved to be ultimately uniformly bounded.Also,the influence of the proxy mass is studied,by which the security of the system is improved.Finally,the simulation and experimental results show that the adaptive proxy-based robust control can realize trajectory tracking and is robust to distur-bances/uncertainties.Also,the proxy mass can make the trajectory of the PMAs-driven exoskeleton present an over-damped response,which effectively improves the safety of the system.In the aspect of active training of the PMAs-driven exoskeleton,this paper analyzes the stiffness characteristics and compliance description of a joint,and proposes a compliance control method for the PMAs-driven exoskeleton based on the antagonistic configuration of PMAs.Firstly,the adap-tive updating law of an echo state network(ESN)is constructed,and the ESN-based super-twisting control(STC)is proposed,which ensures the asymptotic stability of the closed-loop system.This method not only solves the chattering problem of traditional sliding mode control but also effec-tively improves the robustness of the system.Then,according to the compliance description of the joint,a compliance controller of the PMAs-driven exoskeleton system is realized by adjusting the nominal air pressure of the antagonistic PMAs.Finally,the simulation and experimental results show that the proposed method can realize the trajectory tracking of the PMAs-driven exoskeleton and adjust the compliance of the joint,simultaneously.