Design And Implementation Of Soft Exosuit System Based On Multi-mode Recognition

With the acceleration of the exoskeleton commercialization process,the problem of traditional rigid exoskeleton is cumbersome,difficult to wear,poor comfort,and compatibility with the human body which are becoming more and more prominent.Since the ultimate goal of the exoskeleton is to assist the human body in daily walking,how to effectively identify multiple modes and assist in multiple environments has become the focus of exoskeleton researchers.As a kind of exoskeleton,the soft exosuit is lighter,easier to wear,and more compatible.Compared with the traditional exoskeleton,soft exosuit is more suitable for gait assistance.However,the existing soft exosuit adopts multi-joint coupling assistance,which makes the wearable structure complex and does not have universal applicability.The driving device is too high in mass and extra load,which cannot be carried for a long time.The power mode is single,without the ability of multi-mode recognition and multi-environment power.Based on the soft exosuit and mode recognition algorithm,this paper makes an in-depth study on the strategy of assisting the human body to walk.In order to solve the problem of poor applicability of coupling assistance,proposing a multi-environment ankle-assistance strategy.By analyzing the physiological structure and motion mechanism of each joint of the lower limbs,establishing the lower limb model,and the biomechanical data of each joint of the lower limbs are collected and compared under three environments.Considering that in the daily environment,the level ground accounts for a relatively large proportion and the ankle joint’s plantarflexion flexion takes a relatively high proportion of work,the comprehensive analysis shows that the multi-environment ankleassistance benefits the most.To solve the problems of poor universality of soft exosuit structure and high quality of drive device,designing an integrated soft exosuit structure and its portable drive box.The design of soft exosuit based on the physiological structure of lower limbs is applicable to a variety of modes.The exosuit of lower limbs is integrated into shoes to realize integrated wearing,reduce the wearing complexity and improve the applicability.The traditional belttype driving device is improved to be a portable driving box with a backpack and belts,which can enhance the stability in multi-environment.The overall weight of the portable driving box is 2.8kg,which is lighter than most of the soft exosuits.Aiming at the problem of hysteresis in traditional mode recognition,proposing a mode recognition method combining vision and soft exosuit.Multi-threading is used to collect three-dimensional point cloud and three-dimensional euler Angle.The euler Angle is used to correct the point cloud coordinates,and a recognition model of CNN+HMM+QUE is designed.Based on the recognition model,the movement mode is judged in real time.A position control strategy based on feedback force is proposed,which transforms the force control into position control through the system stiffness,simplifies the control model,and optimizes the power curve through the feedback force to improve the power effect.In order to evaluate the performance of the multi-mode recognition soft exosuit system,a mdoe recognition and assistance control is carried out.A multi-mode recognition and multienvironment assistant experimental platform is built,and the performance of the recognition algorithm is evaluated by using the identification accuracy and identification delay.The results show that the algorithm could identify the motion mode in real time,and the identification accuracy could reach 98.7%.The power and torque of the ankle joint under the condition of flexible exoskeleton assisted and unassisted are compared in three environments,and the results show that the soft exosuit can assist.Finally,the paper summarizes the whole work and looks forward to the future research direction.