Research On Compliance Control Technology Of Lower Limb Exoskeleton Robot

With the deepening application of lower limb exoskeleton robot in rehabilitation training,motion protection and enhanced action ability,problems such as unclear individual motion feature recognition,poor motion flexibility and poor human-computer interaction coordination gradually become prominent in the process of extensive application of lower limb exoskeleton.In this paper,the recognition ability of lower limb exoskeleton to human motion state,flexibility of joint motion and coordination of control system are studied.The motion capture system is used to analyze typical movements in daily life,extract joint movement rules in human movement,as a precursor condition for the identification of motion intention,and solve the problem of unclear mapping relationship between individual motion characteristics and motion state in the process of human movement.In this paper,human motion data acquisition is carried out based on IMU sensor.Kalman filter method is used to improve the accuracy of IMU sensor data acquisition.The lower limb motion state signal acquisition system was established by dynamic extraction of the current lower limb motion Angle characteristics combined with the sliding window.A motion state recognition method based on DTW algorithm is designed.Combining real-time motion state acquisition and motion state recognition method,the real-time dynamic recognition function of lower limb motion features is realized,and a method to solve the motion intention recognition is proposed.Combined with the physiological characteristics of human body,the structure design of the lower limb exoskeleton was carried out,and the kinematics,dynamics and ZMP stability analysis of the exoskeleton model were carried out.The results show that the structure of the lower limb exoskeleton robot designed in this paper has certain ability of self-stability and complex motion,which provides a structural basis for the research of the compliance control algorithm of the lower limb exoskeleton in complex scenes.In terms of control,a joint variable stiffness compliance control algorithm is proposed.Combining the dynamic characteristics requirements of rapid response and accurate tracking of lower limb joints,a fuzzy PID position control optimization design is carried out.By analyzing the coupling relationship between human-computer interaction forces and the contact relationship between the lower limb exoskeleton and the external environment under different operating conditions,the optimal design of variable stiffness control is carried out,and finally the adaptive variable stiffness adjustment of joints is realized,and the track tracking ability and flexibility of single joints are improved.Combined with the motion state acquisition and motion recognition system and the joint variable stiffness compliance control algorithm,a closed loop control system of lower limb exoskeleton based on active and passive dual-mode control strategy is designed.The simulation platform of the lower limb exoskeleton robot is built to verify the operation effect of the above algorithm and control system in the overall closed-loop system.The results show that the system can adaptively select the control strategy according to the use scenario,can accurately identify the human motion intention,and has good human-computer interaction ability.Finally,the closed-loop control system of the active and passive dual-mode control strategy is arranged on the lower limb exoskeleton prototype to carry out various tests,verify the track tracking ability and external force compliance of each joint,verify the overall effect of the closed-loop control of the system,and improve the overall coordination and compliance of the lower limb exoskeleton robot.