Research Of Body Size Adaptive Gait For Lower Limb Exoskeleton For Hemiplegic Rehabilitation

Hemiplegic rehabilitation Lower extremity exoskeleton is an exoskeleton robot that assists hemiplegic patients in walking training through machine motion and assists them in gait rehabilitation.Gait planning is an important part of exoskeleton robot,which determines the trajectory of exoskeleton and affects the training effect of gait rehabilitation.At present,most of hemiplegic exoskeleton gait planning does not consider the difference of people,and adopts a fixed gait trajectory,and the difference of people is mainly reflected in body size.People with different body sizes have very different gait trajectories during walking.Therefore,how to plan the gait trajectory to match the wearer's body size is the main goal of this article.For patients with hemiplegia in different periods,the exoskeleton adopts two-leg power and one-leg power modes.For these two different modes,this paper proposes two-leg and single-leg gait planning models,which can plan the corresponding gait trajectory according to the wearer's body size parameters.Firstly,in the two-leg dynamic mode,this paper proposes a gait planning model for legs based on body size,and uses the dynamic motion primitive(DMP)method to establish the basis of the gait trajectory model of the hip and knee joints.After collecting the normal people's gait trajectory,the local weighted regression method is used to fit the parameters of the dynamic motion primitive forced term,and the neural network method is used to fit the step length and cycle parameters suitable for patients with 11 body sizes such as height,weight,and leg length.After inputting the DMP gait trajectory model,combined with the forced term parameters,the gait trajectory of the hip and knee joints of hemiplegic patients suitable for a specific body shape can be obtained.In the enperiment,the comparison errors of the step size,period and actual data obtained by the three normal human body parameters and the two-legged model at different speeds are 0.03-0.05 m and 0.02-0.04 s,and can be planned according to the predicted parameters.The gait trajectory similar to the original trajectory shows that the method can plan the targeted gait trajectory according to the patient's body size.Secondly,for the one-leg dynamic model,this paper proposes an one-leg gait planning model based on body sizes.The single-leg model adopts the method of gradient descent tree.After collecting the joint motion information of one side of the normal person as input,and the joint motion information of the other leg as a standard,the single-leg model is trained.At the input end of the model,the angle and angular velocity of the hip and knee joints of the healthy leg at the time of the current movement are collected at a frequency of 120 HZ through the joint angle acquisition system.After Kalman filtering,the single-leg model is input to make predictions,and the current time can be predicted.The motion information of the hip joint and knee joint of the affected leg is input into the exoskeleton driving device,and the movement trajectory of the affected leg can be obtained.In the experiment,after verifying the data of the left and right legs of three normal people,the average error of the angle between the gait trajectory obtained from the single-leg model and the actual trajectory at different speeds is 1.15 degrees to 2.87 degrees,which improves the adaptability and difference of the exoskeleton.This paper proposes a gait planning model based on body size features for two exoskeleton modes.Compared with traditional gait planning methods,it can match more suitable gait trajectories according to the body size of different wearers,and can be used for early and middle and late stages.Of hemiplegic patients provide better gait rehabilitation training effect.