The method of structure configuration for vehicle leaf-spring based upon BP Neural Network

Considering that vehicle leaf-spring has the typical characteristics of combination, a method of structure configuration is put forward in configuration design for vehicle leaf-spring. The main problem of structure configuration is described aiming to the feature of configuration design for vehicle leaf-spring. Through the generalization of structure type of leaf-spring and the corresponding design requirement, output and input of Neural Network is built and coded in binary. Meantime, in order to decrease the number of sample and simplify the process of sample train and the structure of network, two BP Neural Networks are built to solve the structure features of cross section & section and leaf end & spring eye respectively, which are trained by helps of Neural Network tool in Matlab 7.1. The result of simulation proves that the BP model can meet the demand of solving structure configuration for vehicle leaf-spring.