Mechanism Design And Cooperative Control Of Upper Limb Rehabilitation Robot Based On Motion Intention

Millions of people are worldwide suffered from stroke which leads to dysfunctions of motion.Medical researches are shown possibilities for these patients to regain their motor functions through timely,repetitive and task specific rehabilitation training.Thus,numerous exoskeletons are developed for limbs,and most of exoskeletons are used for the rehabilitation training of patients with dyskinesia of the upper limbs.The exoskeletons usually have ergonomic structure.In addition,the exoskeletons are ability to identify the movement intention of human upper limbs.Thus,the exoskeletons can provide a safe and comfortable rehabilitation training environment for the patients.In this paper,the gravity-balanced bionic upper limb exoskeleton is designed by comprehensively considering the physiological structure.A mapping model of humanmachine interaction information and motion intention recognition is established to estimate human motion intention accurately.Then,the human-machine interaction controller is constructed to realize the interaction movement between the wearer and the upper limb exoskeleton.Based on the common problems existing in the current research of upper limb exoskeleton,three contributions are focused in this paper: the structural design of upper limb exoskeleton,the recognition of intention of the upper limb motion of and humanmachine interaction controller.The specific research content includes the following aspects:A gravity-balanced bionic upper limb exoskeleton based on spring assistance is designed.The forearm structure of exoskeleton is bionic as forearm bones structure of human in order to tolerate slight internal rotation and external rotation during elbow rehabilitation.In addition,the gravity balancing technique of spring assistive mechanisms is analyzed based on the stiffness matrix analysis.The spring assistive mechanisms comprises two ideal zero-free-length springs directly installed to the exoskeleton without using any auxiliary link.In addition,the finite element analysis is used to verify stress of exoskeleton.Furthermore,the rationality of the structure of the upper limb exoskeleton is analyzed by ADAMS,the function and reliability of the upper limb exoskeleton system are tested.Moreover,the Lagrange theorem is used to establish the dynamic model of the gravity-balanced bionic upper limb exoskeleton.A novel closed-loop model based on surface electromyography(s EMG)comprised a long short term memory(LSTM)network and discrete-time zeroing neural algorithm called zeroing neural network,which is developed to estimate joint angles and angular velocities of human upper limb with joint damping.The dynamic model of human upper limb with joint damping is set up as the initial equation.Then,the LSTM network is proposed as an open-loop model which described the input-output relationship between the s EMG signals and joint motion intention.Besides,a novel closed-loop model is built via zeroing finding method for eliminating the predicted error of open-loop model and improving the accuracy of motion intention recognition.Founded on the s EMG signals,the continuous movement of human upper limb joint can be successfully estimated via the novel closed-loop model.The results show that for simple joint movements,the closed-loop model is used to estimate the movement intention of human upper limb with high accuracy.In addition,the clinical experiment data of movement intention is analyzed to verify that the closed-loop model can effectively identify the movement intention of patients with upper limb motor dysfunction.A human-machine interaction controller based on zeroing neural dynamic model is proposed to guarantee system stability and convergence.In addition,an iterative learning controller with zeroing neural dynamic model and disturbance observer is proposed to solve the conflicting of human-machine interaction and disturbances in system.This paper presents a theoretical framework,which is used to process rigorous stability analysis of human-machine interaction control.Combining tracking error dependent weight vector with zeroing neural algorithm is aimed to reduce the conflicts in various operations of motions between upper limb and machine,in which zeroing neural algorithm has strong robustness and noise suppression for solving time-varying nonlinear problems.Moreover,the disturbance observer is used to deal with eliminating the disturbance.Along with these factors,flexibility is also considered for the proposed controller which allows parameter adjustment for operator.In addition,simulations of the controller can effectively and safety assist upper limb movement in different training stages.A prototype of the gravity-balanced bionic upper limb exoskeleton is fabricated,and preliminary tests on subject verified usability of the gravity-balanced bionic upper limb exoskeleton.In addition,the exoskeleton is experimented based on the proportion differential controller without subject.Moreover,the human-machine interaction controller is used to proceed rehabilitation training.The experimental results show that the humanmachine interaction controller designed in this paper has speedy tracking performance,furthermore,the rehabilitation training in different modes is achieved by the upper limb exoskeleton.Moreover,the performance of upper limb rehabilitation exoskeleton is evaluated to verify the effect of upper limb rehabilitation training.Based on the above research,the motion intention of human upper limb is recognized with high accuracy.In addition,a human-machine interaction controller is used to guarantee system stability and convergence.Thus,the human-machine interaction controller is improved the practicability and applicability of the upper limb rehabilitation exoskeleton which realizes the cooperative movement between the rehabilitation exoskeleton and the subject.