Research On Human Gait Perception Method Of Lower Limb Exoskeleton Robot

Wearable lower limb booster exoskeleton robots can effectively enhance the ability of humans to march long distances under heavy loads.It combines the advantages of mechanical structure and human wisdom,integrates them perfectly,and maximizes the advantages according to human needs.Wearable lower extremity booster exoskeleton robot system is a comprehensive product of multiple disciplines,which is one of the most popular research directions at present.This paper focuses on the research and design of the perception system of wearable lower extremity exoskeleton robots,establishes the theoretical model of human kinematics under various motion states,analyses human kinematics,establishes the perception hardware platform,and studies and verifies human gait recognition algorithm and human gait prediction algorithm.According to the theory of human limb kinematics and human anatomy,the method of acquiring and analyzing the motion parameters of lower limbs under the normal motion state of human body based on the attitude analysis system of high-speed camera is studied.This paper mainly studies the human motion characteristics in four processes: normal walking,sitting and standing,squatting and standing,and ascending and descending stairs.Midas software is used to acquire the human motion data presented by four human motion states,and the data are analyzed by using MATLAB simulation software.Using digital filtering technology,the angular curves of back inclination,hip joint,knee joint and angular acceleration of human body under the above four motion states are obtained.Combining with the human motion characteristics,we extract the human motion characteristics data in four motion states and use them as sample data.According to the characteristics of the data on the time axis,the gait of human normal movement is divided.This paper studies and analyzes the types of motion feedback information of human lower limbs,establishes the motion feedback signal collected by the wearable lower limb assisted exoskeleton robot sensing system,determines the type of sensing signal,and then determines the sensor selection.Combining with the mechanim of human lower limb movement,the layout of sensor module,the design of hardware circuit,the establishment of multi-sensor signal acquisition circuit,the analysis of human motion state transition characteristics,the study of the perception algorithm to determine human motion intention,can determine the wearer's current posture and standing intention when sitting and standing,walking intention when standing,and step intention and stop during walking.Intentions,etc.Based on the existing gait data,the gait prediction for the next moment is completed.Finally,human gait perception algorithm and gait prediction algorithm are validated on the lower limb exoskeleton robot perception hardware platform,and on this basis,the algorithm is modified and optimized.The experimental analysis shows that the improved human gait perception algorithm based on fuzzy-proportional perception perceives 3600 gait data based on plantar pressure,with a correct rate of 93.3%.The improved least squares support vector regression based on time series predicts the human gait data at the next moment,with a correct rate of 96.6%.