Research On The Method Of Wearer-exoskeleton Human-robot Interaction

The lower extremity walking and rehabilitation exoskeleton robots for the reconstruction of motor function of paraplegic patients have gradually gained widespread attention in the society.Existing exoskeleton robots can only complete the gait walking task in a specific environment by acquiring the simple motion intention of the human body,and cannot meet the needs of paraplegic patients to walk independently and safely while wearing walking exoskeletons.Therefore,it is necessary to realize the ”human-in-the-loop”human-computer collaboration mechanism through the human-computer two-way interaction technology to improve the safety and efficiency of the human-computer system during the walking process.To this end,this article establishes a two-way interaction of the wearer-exoskeleton system with a human-machine control path and a machine-tohuman feedback path to improve the adaptability of the wearer and exoskeleton.The main contents of this article are as follows:(1)Human intention-behavior mapping based on EMG: According to the characteristics of the mutual transfer between the human EMG signal and the movement during the wearing exoskeleton walking,the non-negative matrix decomposition is used to select the appropriate upper limb muscle as the human-computer interaction interface,the SVM is used as the classifier,and the algorithm combined with the state transition matrix Recognition of the current movement intention improves the accuracy of system classification,and the recognition rate of human intention accurately reaches 93(2)Realize the feedback of the movement state of the exoskeleton based on the sense of touch: use vibration feedback to establish the feedback path from the exoskeleton to the human.It is designed and implemented that four vibrators are combined to form different vibration modes to provide different vibration feedback to prompt the patient's exoskeleton to perform the motion and feed back the status information during the exoskeleton motion to the wearer.After experimental analysis,vibration feedback can help the wearer to better use the exoskeleton robot.(3)Construct a wearer-exoskeleton human-computer interaction mode: use the humanin-loop method to design a wearer-exoskeleton human-machine interaction mode,and optimize the given human-computer interaction objective function by adjusting the interaction parameters.This mode can reduce the difficulty for the wearer to use the exoskeleton robot,change the active application process to the human-machine adaptation process,and help the wearer to become familiar with and use the exoskeleton robot faster.This paper designs a ”human-in-loop” human-computer bidirectional interaction mode that will enable the wearer to accurately control the exoskeleton and perceive the exoskeleton motion state,form a wearer-exoskeleton human-computer interaction loop,and circulate the information between the two to connect the wearer The unidirectional adaptation process of actively adapting to the exoskeleton becomes that both the exoskeleton and the wearer adapt to each other.The man-machine two-way interaction mode can introduce the changes of the wearer during the training process to the man-machine system,making the man-machine system more intelligent and safe.