Modeling And Application Of Lower Extremity Exoskeleton Dynamics Based On Human Gait Data

The lower extremity exoskeleton robot is a wearable device that combines the human body mechanism and mechanical characteristics,and has been widely used in rehabilitation,military,earthquake relief and other fields.With the increasing demand for rehabilitation therapy,the research on lower extremity exoskeleton robots has attracted more and more attention.Lower extremity exoskeleton robots can help patients with spinal cord injury and central nervous system diseases caused by lower limb dyskinesia.The lower extremity exoskeleton robot can provide active and passive rehabilitation training modes for patients.For patients with lower limb paralysis,the lower extremity exoskeleton robot can drive the patient to perform passive rehabilitation training,which plays a key role in the early rehabilitation of the patient;In patients with loss of exercise capacity,the lower extremity exoskeleton robot provides active training for patients and increases the initiative of patients to participate in rehabilitation training.How to realize human-like smoothing control to make training comfortable and effective is still the hotspot of research and popularization at present,and the key points to achieve comfortable and effective smooth control are mainly focused on human-like identification and control strategies.Research on the modeling and control aspects of robots has been carried out.The main research contents are as follows:(1)Analysis of the lower extremity exoskeleton robot system is carried out.For the existing lower extremity exoskeleton system,detailed analysis is carried out from five aspects: mechanical system,sensor system,control system,drive system and energy system.(2)Analysis of the law of lower limb movement.Based on the theory of anthropometry,the law of human walking is studied.The motion trajectory of the lower limbs is measured by the motion capture system on the established motion capture site.The motion of the lower limbs is analyzed by solving the trajectory of each joint center point.(3)The dynamic modeling of the lower extremity exoskeleton is completed.By simulating the robot model,the Lagrange equation is used to derive the multi-input,multi-output and high-coupling nonlinear dynamics models of the two-legged two-link form of the lower extremity exoskeleton robot,and the dynamic characteristics are analyzed in detail.(4)Designed a lower extremity exoskeleton robot control system.In order to realize the result effect that the lower extremity exoskeleton robot can track the gait trajectory of normal people smoothly and quickly,PD control,adaptive control and robust adaptive PD controller are designed respectively.Among them,the PD control can realize the fixed-point control of the lower extremity exoskeleton robot while ignoring the system gravity and disturbance,and realize zero tracking error of position and speed.According to the uncertainty of the dynamic model of the lower extremity exoskeleton robot,the adaptive control system is designed to achieve the tracking control effect.Considering the disturbance between the system and the outside world,the robust adaptive PD control system is designed to realize the fast tracking of joint angle and angular velocity under the condition that the error and disturbance upper bound are unknown.(5)Verification of the effectiveness of applying a human-like modeling method to identify a dynamic model of a lower extremity exoskeleton robot.Through the reverse simulation method,the trajectory of human walking is used as the desired output of the lower limb exoskeleton,and the joint torque is used as the main control input signal during walking,the simulation of the control effect in the presence of model error is completed,and the validity of the proposed method is verified.