Research On Enabled Technologies Of The Human-robot Interaction For Assistive Exoskeletons

Exoskeletons have excellent application prospects in the fields of power augmentation,health assistance,rehabilitation and high-tech manufacturing.Essentially,exoskeleton is a kind of tool that centers on human beings,guarantees human health and gains human trust.Over the past 20 years,the research on exoskeletons has increased exponentially.However,the large-scale application of exoskeleton in various fields has not yet been realized.The fundamental reason is that the core characteristics of exoskeleton as a "tool" have not been perfected.There are some problems in human-exoskeleton interactions,such as the low stability of human-robot interaction,the limitation of human-robot connection on human movement,and the deviation between the human and exoskeletons.The interactive forces generated by the misalignment threaten human safety,dislocation of exoskeleton power supply time,dislocation of strength,interaction lag between human and computer,and problems such as large volume,self-weight and inconvenience of on-off.In order to achieve the macro goal of the usability of exoskeleton ergonomics,this paper studies the key technologies of exoskeleton human-computer interaction from the following five aspects.1.The key technologies of physical human-Robot Interaction(p HRI)and physical humanRobot interface(p HRi).The uncontrollable interaction forces caused by the deviation of humanexoskeleton dimensions and axis interfere with and restrict the human motion.With the help of exoskeleton,people need to make more efforts and consume more energy,which hinders the human-exoskeleton applicability and safety of exoskeleton.Based on a comprehensive study of human bio-anatomical characteristics and physical characteristics of exoskeleton engineering,combined with the advantages of current exoskeleton bionic design: concise structure,no motion singularity,no interference with human beings,the human musculoskeletal model is described as "simplified human model + human-machine deviation model",as exoskeleton design reference.Combining with the design method of self-aligning mechanism,this paper designs an adaptive lower limb rehabilitation exoskeleton mechanism and human-exoskeleton connecting device.The increased degree of self-adaptive freedom makes the exoskeleton motion space closer to the human natural motion space.2.Key technologies of human-robot interaction(c HRI)and human-robot interface(c HRi)for exoskeleton control.In order to drive exoskeleton accurately at joint level and time point,this paper describes human natural motion as "simplified human motion + human-robot deviation motion",and divides human joint variables into main joint variables and deviation variables.In order to make the exoskeleton movement cover the natural movement of human body,the exoskeleton joint variables are divided into control variables and adaptive variables accordingly.In the closed motion of man-machine system,the D-H 4*4 matrix transformation is used to obtain the activating and adaptive rules of exoskeleton at the joint level,so as to realize the precise activating and control of exoskeleton at the joint level.Taking the rehabilitation exoskeleton of lower limbs as an example,this paper calculates the driving rules of decoupled exoskeleton of the hip in three directions: flexion/extension,abduction/adduction and external rotation/internal rotation.The theoretical method makes the human-exoskeleton deviation within the controllable compensation range of the movement of the exoskeleton mechanism,and makes the use of the adaptive redundancy degree of freedom not only empirical and supplementary,but also precisely controllable.3.The human-robot interaction impedance control rule based on mathematical model regards human body as mechanical impedance body.The human-exoskeleton system constitutes a flexible adaptive system of impedance and admittance.In this paper,a human-robot impedance control strategy for unilateral adaptive exoskeleton rehabilitation of lower limbs is proposed.Human-robot impedance control is a typical feedback control with time delay and inaccuracy of inertia compensation.In this paper,we discuss the relationship between human motion intention and muscle cooperation based on load-assisted exoskeleton experiments,and the adaptive characteristics of neuromuscular control strategy in external force,so as to achieve muscle coordination.Quantitative analysis of contraction index(CCI)and muscle synergy analysis module(MSA)reflects the adaptive ability of human muscle activity under the assistance of exoskeleton,and obtains the theoretical basis of human-computer interaction control of exoskeleton EMG feed forward.4.Based on the observation of human autonomous optimization ability,the humanexoskeleton interaction system is optimized.The performance of human autonomous optimization is embodied in different aspects,such as visual gait adjustment,deeper neuromuscular signals,synergistic muscle contraction and muscle cooperation.This paper draws inspiration from the changes of human muscle activity in exoskeleton human-computer interaction experiments,and proposes an optimization model.The control strategy of the exoskeleton device can be changed systematically to adapt to the optimum performance of human to the greatest extent.EMG signals in human interaction were measured,and the control parameters of exoskeleton joints were determined by sampling valid data to reduce the level of joint paired muscle synergistic contraction(CCI).During the optimization process,EMG signals are continuously measured,and the exoskeleton periodically changes the human-computer interaction mode defined by the control strategy.Under each control strategy,the CCI level of the stationary state matches a power-driven model.After all the control strategies are matched,a new iteration is formed and the next generation control strategy is calculated by covariance matrix.5.Phased evaluation method of exoskeleton is proposed on the basis of literature research.Each stage includes objective performance matrix and subjective performance matrix.In this paper,a controlled laboratory test and evaluation is carried out for an augmentation exoskeleton prototype and the auxiliary assembly exoskeleton prototype.The objective evaluation tools used in the evaluation include electromyogram signal,time-space parameter of gait and human-robot contact pressure test.Subjective assessment tools are NASA-TLX subjective performance matrix,to assess the overall psychological load from six dimensions: mental demand,physical demand,time demand,performance,efforts and frustration level.The evaluation results of augmentation exoskeleton show that it has excellent loading capacity,while the humanexoskeleton interaction follow-up performance is limited by the excessive flexibility of humancomputer connection.The activities of knee power muscle decrease by more than 10%.Subjective performance shows that exoskeleton reduces the physical demand by 65%,but increases the mental demand by 50%.In the evaluation results of assisted assembly exoskeleton,most of the vibration effects of manual tools are absorbed by exoskeleton.The muscle activity of workers is reduced by 20% on average,and the human-machine contact pressure is reduced by 20.4%.