Nonlinear dynamics model of variable gap magnetorheological buffer based on generalized regression neural network
The kinetic behavior of variable-gap magnetorheological buffers has extremely nonlinear characteristics,and the establishment of an accurate kinetic model is the basis for accurately controlling the output characteristics of the buffer.To this end,a nonlinear dynamics model based on generalized regression neural network(GRNN)is proposed.Firstly,a prototype variable-gap magnetorheological buffer was fabricated and an impact test platform was built,and the impact test data under different working conditions were collected as the buffer dynamics database;then,the performance of the GRNN model was verified and compared with the back propagation(BP)neural network model,and the results showed that the GRNN model could effectively map the nonlinear dynamic performance of the buffer and its accuracy was higher than that of the BP neural network.The results show that the GRNN model can effectively map the nonlinear dynamic performance of the buffer and its accuracy is higher than that of the BP model;finally,the accuracy of the GRNN model in predicting the output of the buffer under the unknown operating conditions is verified,and the relative errors of the peak buffer force,maximum displacement,maximum velocity and the experimental results are 4.71%-10.2%,0.32%-4.18%,and 0.097%-2.5%,which indicates that the GRNN model can accurately predict the output characteristics of the buffer under the unknown operating conditions.
NETL
NSTL
万方数据
