Research On Structural Design And Control Strategy Of Muiti-dof Lower Limb Exoskeleton Rehabilitation Robot

In recent years,the problem of population aging in the world has become more and more serious,resulting in an increase in the incidence of hemiplegia among the elderly.Coupled with factors such as traffic accidents and labor injuries,the number of people with loss of lower extremity motor function has gradually increased.The lower limbs,including the hip,knee and ankle joints,play an important role in people’s daily life.Because the previous rehabilitation training method is to use professional rehabilitation personnel to conduct targeted training for patients,the contradiction is that there are many patients,and rehabilitation physicians have great deficiencies in both quantity and quality.In addition,rehabilitation medical equipment The function is too single,so there is a need for efficient and well-functioning equipment research internationally.Aiming at the limitations of rehabilitation methods and assisted walking for patients with lower extremity motor dysfunction,a Multi-DOF lower extremity rehabilitation robot was designed,including structural design and control methods.Firstly,the skeletal structure and muscle composition of the lower limbs were analyzed,and the relationship between joint movements and their interactions was sorted out,and a Multi-DOF lower-limb exoskeleton rehabilitation robot was designed to realize the rehabilitation training of the lower limbs with 10 degrees of freedom,including single-limb hip,Knee and ankle joints,the degrees of freedom are distributed as 2,1,and2,which can realize various rehabilitation training and assistance methods such as erection,leg lift,hip abduction,ankle and lumbar lateral deviation,and the design size conforms to the standard of ergonomics.Secondly,theoretical rationality analysis of lower extremity exoskeleton rehabilitation mechanism is carried out.Create the D-H coordinate system of the serial part of the hip and knee joints,solve the forward and inverse kinematics solutions of this part by combining the algebraic method and the improved PSO optimization algorithm,establish the constraint equation of the parallel part of the ankle joint,and prove the correct solution of the mechanism by numerical method.Use MATLAB and ADAMS to model the mechanism and simulate the kinematics to prove the correctness of the mechanism.Thirdly,the surface EMG signals of various movements of the feet are extracted and classified as control instructions.The sEMG signals of various motions of the feet are extracted by Delsys equipment,and the signals are preprocessed to obtain the sEMG signal features in the time domain and combine their feature vectors.The data is classified by the three classification algorithms of LDA,RNN and CNN,the input motion is identified and classified,the classification effect of different algorithms is compared,and the CNN algorithm with the best classification effect is selected as the classification target algorithm as the follow-up control system research.Finally,the structural innovation of each part based on the basic experimental platform and the individual control experiments of the ankle joint are introduced.Through the wearing experiment experience of the basic experimental platform,three innovative structures are specifically proposed,and the control system platform based on the embedded development platform is designed as a separate control system for the ankle joint mechanism.Through the ankle joint feedback experiment,it is proved that the The rationality of some institutions and the direction of follow-up research.