Design And Integration Of Follow-up Exoskeleton Robot For Rehabilitation

With the aging of society becoming more and more prominent and the number of disabled people is increasing year by year,in order to meet the needs of helping the elderly and helping the disabled,the exoskeleton rehabilitation robot came into being.The exoskeleton rehabilitation robot is a kind of electromechanical intelligent equipment with high imitation and high coupling.Based on structural design,gait planning,control algorithm,and so on,a set of follow-up exoskeleton robot for rehabilitation is developed,which combines theoretical analysis and experimental verification.The adaptive controller and based on Openpose human-inspired algorithm were designed.The main findings of this dissertation are summarized as follows:(1)A set of follow-up exoskeleton robot for rehabilitation is designed.which is modeled and optimized by Solidworks.The system includes the following training frame and exoskeleton robot for rehabilitation.Learn the physiological structure of human body and determine the degree of freedom configuration.The training frame provides the patient with the auxiliary protection function of servo at all times,ensures the stable and safe operation,and provides efficient and reliable human-robot interaction.(2)Kinematics analysis and gait planning of exoskeleton robot for rehabilitation.A forward kinematics models of exoskeleton robots were established by D-H coordinate method.To analysis of human gait,based on cubic spline interpolation method,five key interpolation nodes were identified.Gait planning of hip and ankle was carried out in the initial stage of gait and continuous walking stage,respectively.Kinematics knowledge was used to solve the motion trajectory of knee joint.Finally.MATLAB was used to draw the desired trajectory map of joint.(3)Exoskeleton dynamics modeling and adaptive control algorithm design.The extoskeleton dynamics model is established by lagrange function method.The unknown parameters in the human-robot system dynamics equation are estimated based on neural network approximation algorithm,and the unknown time-varying periodic disturbances are compensated.The stability of the designed controller is verified by Lyapunov function direct method.(4)Design an Openpose-based human-inspired control algorithm and complete the experimental verification.Based on the Openpose algorithm,we use binocular camera to identify the wearer's shoulder skeleton,and design the human-inspired control algorithm to control the relative distance between the follower and the wearer in the safe range,so that the follower can follow the wearer for rehabilitation training.Finally,the comparison experiment between the design and PD control shows that the control algorithm designed in this paper can reproduce the normal human gait,improve the rehabilitation efficiency of patients,and verify the realization of the rehabilitation function of the platform and good human-robot interaction.