Estimation And Application Of Interaction Force Between Human And Lower Limb Exoskeleton Device During Locomotion

The interaction force estimation plays a paramount role in human-robot coupling system since it can not only provide valuable clues for the structure design and optimization of the wearable robotic devices,but also promote the understanding of the human-robot interaction mechanism.However,due to the current immature force sensing technology,the complexity and diversity of interaction force,there exists significant challenges in directly measuring interaction forces.Developing force estimation methods under simulation environment can provide important theoretical and technical supports for the actual measurement of the interaction force.To this end,with the help of the OpenSim simulation platform,this paper focuses on the estimation of human-exoskeleton interaction force during locomotion.Firstly,the positions of the markers are optimized by a proposed bilevel optimization method in order to obtain a better-matched virtual kinematic model.Then,based on the joint angle information of the optimized virtual model,an interaction force estimation model satisfying Hooke’s law is established.Afterwards,based on Solid Works,a humanoid exoskeleton robot is constructed,and the exoskeleton model is driven by forward dynamics,combined with the proposed interaction force estimation model,to estimate the interaction force between human and exoskeleton during gait process.Finally,based on the proposed interaction force estimation method,the optimization problem of the exoskeleton strap’s location in the gait process is modeled,and the established optimization problem is solved from the single-objective and multi-objective perspectives,respectively.The main works of this thesis are summarized as follows:(1)A bilevel optimization algorithm based on Inverse Kinematics is firstly proposed.In order to improve the matching degree between the virtual model and the real model,based on the Pattern Search Algorithm(PSA)and the IK solving tool in OpenSim,a bilevel optimization method is proposed to optimize the location of the markers regarding to the virtual model.Through the numerical simulations over six subjects,the effectiveness of the proposed bilevel optimization method is verified.(2)A method for estimating the interaction force of human-robot coupling system is proposed.Firstly,combined with the joint kinematics of the optimized virtual model via the bilevel optimization approach,a model for estimating the interaction force of the human-robot coupling system is established.The exoskeleton model created in Solid Works is then coupled to OpenSim and simulated in Forward Dynamics to determine the spring elastic coefficient in the interaction force model in order to estimate the exoskeleton hip and knee interaction forces.Finally,numerical simulation is given to verify the effectiveness of the proposed interaction force estimation method.(3)An exoskeleton strap position optimization method based on the interaction force model is proposed.Firstly,based on the proposed interaction force model,a mathematical model of exoskeleton strap position optimization problem is established.Then,the established optimization problem is solved from the aspects of single-objective and multi-objective optimization,respectively.In the single-objective solution,the analytical expression of the weight coefficient ratio and the exoskeleton strap position in the optimized objective function is obtained by using the curve fitting method.In the process of multi-objective optimization,a multi-objective particle swarm optimization(MOPSO)algorithm is proposed to solve the established optimization problem,and the Pareto non-inferior solution set meeting the optimization requirements is obtained.Finally,the effectiveness of the proposed method is verified by numerical simulation,and the pros and cons of single-objective optimization solution and multi-objective optimization solution method are analyzed and explained.The results of this paper show that the IK-based bilevel optimization algorithm can significantly reduce the model error and improve the dynamic accuracy of the model.Also,when the interaction force estimation method is used to estimate the interaction force of the exoskeleton joint,the interaction forces at the exoskeleton joint exhibit similar variation tendencies as those of the real motion of the human joint.Moreover,the optimization of the exoskeleton strap position based on the interaction force can design the optimal strap position according to the preference of the decision maker.Therefore,the simulation study of interaction force estimation can not only expect to lay a theoretical foundation for interaction force estimation in human-robot coupling system,but also provide certain technical supports for the design and optimization of exoskeleton mechanical structure during locomotion.