Research On Transparency Design And System Verification For Walking Assistance Exoskeleton

Currently,there is a lack of effective communication mechanism between exoskeleton and the wearer,which cannot meet the significant requirements of safety,naturalness,and efficiency in human-exoskeleton cooperative walking in daily living environment.Therefore,paraplegic wearers put forward high requirements for the transparency of exoskeleton systems.This dissertation focused on the improving the transparency of exoskeleton robot to enhance the safety,efficiency,and naturalness of human-exoskeleton collaboration.The transparency of the exoskeleton was designed in this dissertation and verified on AIDER exoskeleton robot.The main contributions are as followed.The wearer and exoskeleton lack efficient information communication method.And there is urgent demand of sharing information between human and exoskeleton.In view of the above problems,the transparency framework of human-exoskeleton is proposed in this dissertation.The information transmission mechanism of human-exoskeleton behavior was studied and the transparency framework was constructed.According to the transparency framework,the expression method of transparency of exoskeleton robot is designed.In view of the current paraplegic patients can not perceive the man-machine walking state because of the loss of proprioception,two tactile feedback methods that are strengthchanged vibration and position-changed vibration were proposed to represent the gait information of exoskeleton robot.Evaluation results showed that there was no significant difference in recognition accuracy and response time between the two modes( > 0.05).The subjective questionnaire results showed that the subjects were more inclined to SCV mode( < 0.05).To solve the problem that the wearer cannot predict the behavior intention of exoskeleton,which may result to danger,a transparent expression of human-machine behavior intention is constructed.Two feedforward interaction methods that are speechand music-feedforward modes are designed to express the behavior intention of the exoskeleton.Theoretical results show that there are significant differences in recognition accuracy and response time between the two methods( < 0.05),the speechfeedforward method is more suitable for conveying the behavior intention of the exoskeleton.In order to evaluate the effectiveness and safety of the transparency design of exoskeleton from various aspects,an evaluation method for transparency design is proposed from four aspects: subjective score,behavior characteristics,physiological state and human-exoskeleton efficiency.The validity and safety verification results are as followed.First of all,the human-exoskeleton trust degree of the transparent mode was significantly improved compared with that of the non-transparent mode because there were significant differences in the scores of trust scale( < 0.05).Second,the subjects' walking posture improved significantly because significant difference was in the distribution probability of gaze region( < 0.05).Third,Hb O concentration was close to that of normal walking mode( > 0.05).Forth,human-exoskeleton collaboration efficiency is significantly improved compared with non-transparency mode( < 0.05).To sum up,the transparency design of exoskeleton robot studied in this dissertation solved the problem of lacking efficient communication mechanism between human and exoskeleton,and enhanced human-exoskeleton mutual understanding,and the safety,efficiency,and naturalness of human-exoskeleton collaboration.