Modeling And Simulation Of Lower Limb Soft Exoskeleton Robot System

As a wearable robot,exoskeleton robot is widely used in military equipment,medical aid and civil equipment.However,in the actual use,the exoskeleton robot has obvious defects such as inertia of equipment,kinematic constraints and lack of optimal control method.The soft exoskeleton robot is a wearable device based on textile cloth that provides a conformal,unobtrusive and compliant way of contacting the human body to address the above challenges.Unlike most rigid exoskeletons,soft lower limb exoskeletons can be worn like lightweight clothing,combined with a proximally mounted actuator,it provides strength parallel to the muscles of the lower limbs,increasing walking efficiency.This paper mainly studies the development of software exoskeleton,driving platform,controller and optimization method.The research contents mainly include the following aspects:(1)Based on human gait analysis and kinematics analysis of the software exoskeleton,and with the help of MATLAB and Open Sim platform simulation,the overall design of the software exoskeleton is proposed,and the working principle of the software exoskeleton and the design of joint drive platform are also introduced.For the soft exoskeleton,two kinds of exoskeleton controllers are designed,which are position control based on iterative force and admittance-position controller.The position control based on iterative force can accurately detect the beginning time of hip extension and the peak time and amplitude of the tracking auxiliary profile within the error range.The switching admittance-position controller can track the change of the power curve.(2)In order to explore the man-machine interaction and the choice of driving platform application time,a human subject study was carried out to evaluate the effect of multijoint assist and the time of hip extension assisting.In the multi-joint assisted study,subjects ran on a treadmill at a fixed speed under single joint conditions,resulting in reduction of metabolism.The results showed that by boosting the ankle joint,metabolic expenditure could be reduced.In the impact of hip extension time study,four different booster curves were evaluated at two sets of starting and peak times,with subjects walking on a treadmill at the same speed.The results show that the choice of driving time can affect mechanical dynamics,biological joint dynamics and human metabolic expenditure.The boost curve with an earlier start time and a later peak time provided the maximum mechanical power and achieved a higher metabolic reduction compared to the no-power condition.(3)the development and evaluation of Bayesian optimization methods for identifying optimal control parameters in circular walking are presented.Firstly,the optimization method is evaluated by finding the best step frequency of the target.It is found that the time of single parameter optimization convergence to the optimal step frequency is half of the established gradient descent method,which significantly reduces the total energy consumption of the whole experimental scheme.On this basis,through the optimization of peak and offset time of the soft exoskeleton hip extension assist,the Bayesian optimization and the gradient descent method are compared.In this study,the metabolic value was significantly reduced when compared with that without the soft exoskeleton.The results support the hypothesis that the individual differential control method is more effective than the fixed control strategy based on the specific metabolic distribution of the subjects,including peak,offset time and metabolic status.