Research On Human-machine Cooperative Control Strategy Of Mining Exoskeleton Assisted Robot

In the past decade,exoskeleton technology has made continuous breakthroughs and has a very broad application prospect in the field of coal mining.Exoskeleton assisted robot can effectively reduce the physical consumption of workers,improve work efficiency,and significantly reduce the risk of musculoskeletal disease caused by work.Considering that the coal industry mine has a standard compressed air supply system,it is very convenient to obtain compressed air underground,so the mining exoskeleton robot driven by air pressure has obvious advantages.At the same time,compared with the motor or hydraulic drive,pneumatic muscle has the advantages of good flexibility,light weight,high output power to weight ratio,low price,etc.Therefore,this paper designs a mining exoskeleton power assisted robot based on pneumatic muscle,and on this basis,carries out the research of human-computer cooperative control.Firstly,this paper expounds the background and significance of the topic,summarizes the development status at home and abroad,and puts forward the research objectives and contents.Secondly,the design structure of exoskeleton is introduced,and the dynamic model of upper limb is established.On this basis,sliding mode controller and adaptive robust controller are designed to track the joint angle of exoskeleton.The simulation and experiment of the two algorithms are compared to verify the advantages and disadvantages of the two algorithms.Then,the overall control strategy of humanmachine cooperation of mining exoskeleton robot is carried out.The sensitivity of exoskeleton is improved by gravity compensation of force sensing signal.The fuzzy logic reasoning system optimized by particle swarm optimization is designed.The pressure signal after gravity compensation is used as input to judge the motion intention of human body and quantify the joint angle to predict it.Using the quantified joint angle information to infer the spatial motion information of human body,based on a method of motion path generation which establishes the feature mapping relationship between the position of motion space and the motion path,the motion path of joint angle is planned.Finally,the Kalman filter is used to predict and compensate the hysteresis of the generated motion track caused by the force sensing signal,and the effectiveness is proved by experiments.This thesis has 45 figures,5 tables,85 references.