Human-like Control Of Lower Limb Rehabilitation Robot Based On Human Movement Gait Data

With the aggravation of the aging population and the increase of patients with movement disorders,more and more people need lower limb rehabilitation training,and rehabilitation training needs to gradually correct or improve the walking posture and gait of patients.As is known to all,assisted rehabilitation training by manual and experiential methods is the most effective way for doctors,but the number of rehabilitation doctors is far from enough to meet the needs of patients.In order to reduce the working pressure of doctors and help patients to carry out rehabilitation training as soon as possible,the use of humanoid lower limb rehabilitation robot designed based on the law of human movement has become an effective way to carry out rehabilitation training,which has become a research hotspot.Therefore,how to effectively process the movement mechanism of human body,such as gait data,and apply it to the design and control of lower limb rehabilitation robot becomes the key to realize the design and humanoid of lower limb rehabilitation robot.Therefore,this paper takes the lower limb rehabilitation robot as the research object and carries out the design and research of humanoid control system based on human gait data.The main contents are as follows:(1)Analysis of human motion mechanism.The real gait data of human body were obtained by motion capture system and force measuring platform,and the characteristics of the simulated robot dynamic model were analyzed.Considering that the external interference in the acquisition process and the jitter in the walking process lead to the uneven gait motion curve,the filter processing was carried out firstly,and the Fourier fitting method was used to integrate the data,so as to prepare for the application of gait data to the design of lower limb rehabilitation robot.(2)Dynamics modeling of lower limb rehabilitation robot.By imitating the multi-link robot model,the two-link nonlinear model of the lower limb rehabilitation robot was deduced by using Lagrange equation.Aiming at the uncertainty of model parameters,the structural processing of parameters is carried out.(3)Modeling,Simulation and Control of Human Gait Based on OpenSim.In this paper,the human body model was calibrated by motion capture system,and the Angle and predicted torque of human lower limb joint were obtained by model scaling,inverse kinematics and forward dynamics solution,and the motion law was analyzed.In the control system design,the human motion torque predicted by OpenSim was used as feedforward compensation,and the controller based on gait characteristics was designed to complete the humanoid motion of rehabilitation robot.In the verification of the method,the trajectory of human gait is taken as the desired trajectory,and the motion data is input into OpenSim as the control quantity to conduct the reverse humanoid motion verification.The final results show that the method can not only track the real trajectory of human gait,but also make the initial torque change decrease,and the comfort level increase.(4)Adaptive Control of Model Uncertain Rehabilitation Robot Based on Real Human Gait.In this paper,a human-simulated adaptive controller based on human gait data is designed to solve the problem of uncertain robot dynamics model,and the robust stability of the system is analyzed.The simulation results show that this method not only compensates the uncertain part of the robot model effectively,but also tracks the real human gait effectively,and improves the control precision.