The Study On Gait Reconstruction And Gait Tracking Of Biped Robot With Heterogeneous Legs

Biped Robot with Heterogeneous Legs is a new-style robot, integrating intelligent prosthesis and biped robot study. It can simulate the situation that patients with above knee amputation wear intelligent prosthetic, do a mass of repeatability and variety test study of intelligent prosthesis. BRHL provide an ideal scientific test-bed for study of intelligent prosthesis. The paper gave a brief introduction to the model of BRHL on the basis of recent situation and significance of Biped Robot and intelligent prosthesis, and described the structure of bionic leg and artificial leg. Then we introduced the common simulation platform ADAMS and MATLAB/simmechanics, as well as the model built on the two platforms. We mainly focused on the three parts:prediction of knee joint angle, gait symmetry optimization and joint angle trajectory tracking.One of the key questions needed to be solved in the process of use intelligent prostheses was how to predict the prosthetic knee joint angle based on the healthy limbs motion information. The information of angular velocity of the thighs and shins on both sides was collected when people walking through the inertial sensors. Then PCA was used to analysis the coupling relationships of both sides and these relationships were used to predict the knee-joint angle. Finally further optimization was made by the adopting of RBF network. The results proved the correctness and effectiveness of this scheme. The prediction of knee-joint angle has a great significance to energy efficiency and gait to human nature of the intelligent prosthesis.The difference of the right side and left side knee joint’s structure will bring bad effect on the BRHL’s gait symmetry. At the first, the gait symmetry of normal people was analyzed, and as a reference to estimate the BRHL’s gait symmetry. A scheme coupling genetic algorithm with the polynomial transformation was proposed to solve the poor symmetry of BRHL’s step length. In the ADAMS simulation environment, the right knee joint angle was transformed through a linear equation, then the parameters of the linear equation was optimized by using the genetic algorithm. After several iterations, the symmetry of BRHL’s step length was within the normal range. After the improvement, this program can be used to optimize the symmetry between the intelligent prosthesis and the healthy limb.The error source of the existed experimental platform was researched. Then the simulation model was built in the simmechanics platform, with the error signal. The simulation of the joint angle trajectory tracking was researched and achieved good control result, by using the BP neural network PID control algorithm. This program can be used to reduce the experimental platform’s tracking error caused by the wear, gear clearance and other reasons.