Traditional robots have difficulties in quick response to external disturbance.Based on OpenSim simulation and human upright push-recovery experiment,this thesis commits to investigating the control strategy of human ankle joint during human upright anti-disturbance.The main work done in this thesis are summarized as follows:(1)A human musculoskeletal simulation platform based on OpenSim is established.Firstly,the dynamic model of human body is obtained by mechanical analysis of human body.Then the Hill-type muscle model and mechanical model of human body are studied.The parameters of the simulated musculoskeletal model are adjusted according to the existing musculoskeletal model.The forward dynamics simulation engine is investigated in order to achieve the quiet upright stance of the musculoskeletal model on OpenSim simulation platform.(2)Stretch reflex controller is developed.In order to keep the balance of musculoskeletal model under external disturbances,this thesis studies the mechanism of stretch reflex in human body.Also,this thesis completes the development of a stretch reflex controller based on muscle length and velocity,which is capable of maintaining the model balance in the case where the model subjects to external disturbances.Besides,this thesis conducts the human upright push-recovery experiment,after which simulation and experiment data are compared.The comparison results confirm that the data of simulation coincides well with actual data obtained from experiment of human push-recovery response.(3)The affects of muscle force positive feedback and delayed link in motion system on the stretch reflex are explored in this thesis.On the basis of stretch reflex,the part of positive force feedback of muscle and time delay are analyzed and realized,respectively.The positive force feedback of muscle can improve ankle stiffness and delay time can reduce the stability of model upright.(4)The effect of the feedforward to time-delay reflex control is studied.Feedforward link is added to the muscle reflex controller.The simulation data is compared and analyzed with the experimental data of human upright push-recovery experiment.The results show that feedforward can reduce the impact of time delay on human upright.This thesis investigates the mechanism of human muscle reflex control and the rapid response strategy involved in human interaction with the outside.The obtained results can not only provide insights into how human keeping upright stance,but also can promote the bionic study of the lower-limb robotics. |