Human-mechine Cooperation Control Strategy And Optimization Of Ankle Exoskeleton

Exoskeleton has great potential in improving human locomotion ability,such as assisting the elderly to go up or down the stairs,as well as the rehabilitation of the lower-limb disabled,and so on.From the perspective of ankle biomechanics and muscle mechanics,this thesis investigates the cooperation control between ankle exoskeleton and human under the walking task.The main work done in this thesis includes: the mechanical design of the ankle exoskeleton;the design of control system;and the optimization of control strategy.(1)Design an exoskeleton structure that is highly coupled to the ankle joint of human and mechanical selection.The innovation of this exoskeleton is that a Boden rope divides the exoskeleton into two parts,the exoskeleton part and the motor drive part,which greatly reduces the weight of the wearable part of exoskeleton.(2)An ankle exoskeleton torque control method based on two-level controller is proposed.In the proposed two-level controller,the higher-controller determines the desired torque of the ankle joint exoskeleton through a gait cycle time,andthe lowerlevel controller tracks the desired torque curve that input by the upper-level controller through iterative learning.(3)An optimization of ankle exoskeleton control strategy for “human-in-loop” under the human walking task is proposed.In order to improve the assisting effect of exoskeleton to human,the optimization method takes the average value of the rootmean-square of soleus of the left and right legs in each gait cycle of the wearer as the objective function and optimizes the control parameters of exoskeleton controller by particle swarm optimization(PSO)algorithm.In order to well balance the global and local search abilities of PSO,a self-adaptive updating rule is proposed to adjust the inertia,cognitive and social acceleration parameters of particles in PSO.(4)In order to validate the mechanical design,of the ankle exoskeleton,the designed controller and the “human-in-the-loop” optimization method,this thesis conducts some experiments.The experimental results confirms feasibilities of the mechanical design of the ankle exoskeleton,the designed controller and the “humanin-the-loop” optimization method.From the perspective of bionic,this thesis finishes the design of an ankle exoskeleton which is highly coupled to the human ankle joint.Also,a PSO-based “human-in-the-loop” optimization framework is proposed in order to enhance the cooperation efficiency of the ankle exoskeleton and human.The obtained results of this thesis can promote the study of the exoskeleton robotics under some complicated tasks.