Natural Human-robot-interaction Technology And Method For Soft Exosuit

Soft exosuit is a new kind of wearable robot,which has a wide application prospect in medical,military and industrial fields.It has been proved that can greatly reduce the metabolic consumption of human walking and running.As a complex human-robot interactive system,whcih involves lightweight structure design,human intention recognition,motion analysis and robotic control strategy.However,on the one hand,the large weight of the soft exosuit system makes it difficult for the assist effect to offset the extra metabolic consumption which caused by the weight of the system,and the soft exosuit worn away from the center of gravity will increase the physical instability during walking and running.On the other hand,in order to build a natural human-robot interaction system of the soft exosuit,some connections between human psychological information and the soft exosuit should be focued.In this paper,a more reasonable natural human-robot interaction control strategy of soft exosuit is presented from the system design,environmental information perception and human information perception,so as to reduce human metabolic consumption,muscle fatigue and increase human-computer interaction performance.The specific research contents include:(1)System design and environmental information perception,including lightweight soft exosuit construction design and multi-terrain classification.Firstly,based on the inspired design of human body,the lightweight soft exosuit(only 1.8kg)known at that time was proposed,which can assist hip flexion and reduce the metabolic consumption and muscle fatigue during walking and running.Secondly,the biological moment of lower limbs is different under different terrains,and 91% classification results can be achieved by using support vector machine algorithm for five common terrains.(2)Human information perception,including human emotion analysis,motion intention recognition and motion feature analysis.Firstly,the relationship between different emotions and gait was analyzed and the results of emotion classification were applied to the control of lightweight soft exosuit.Secondly,there are two gait prediction methods which based on neural network and based on mean prediction.Finally,the change curves of the biological hip joint moment at different speeds were analyzed,and different control strategies for the lightweight soft exosuit were formulated according to the theoretical reference.(3)Natural human-robot interaction control strategy research.According to the system design,environmental information perception and human information perception,this paper illustrated natural human-robot interaction control strategy for lightweight soft exosuit under single terrain,different speeds and emotion changes,so as to achieve high compatibility of human-robot interaction.In this paper,we invited 5 subjects to test respiratory metabolism rate and muscle fatigue degree by the lightweight soft exosuit under single terrain,different speeds and emotion changes.It was demonstrated that walking on a single terrain reduced the user’s metabolic consumption rate by 11.52%,and reduced the muscle fatigue of rectus femoris,lateralis femoris and gastrocnemius muscles by 10.7%,40.5% and 5.9%,respectively.Running at 6km/h,8km/h and 10km/h can reduce the metabolic consumption rate of users by 7.79%,4.74% and 6.12%,respectively.In different emotional experiments,it was demonstrated that the lightweight soft exosuit control method based on human emotional changes could reduce the metabolic rate of walking under excitiment,depression and fear emotions by 4.5%,6.7% and 3.7%,respectively.This research provides a new idea for the natural human-robot interaction mechanism of lightweight soft exosuit.