BP神经网络在车辆制动性能预测中的应用

首先建立了车辆制动过程数学模型,利用该数学模型对JS2310农用运输车的制动性能进行了计算机仿真,并与试验结果进行了比较,表明模型是正确的为进一步提高该车辆制动性能预测的精度,引入了神经网络技术使用BP网络对不同条件下的多次仿真结果进行第一步学习,将数学模型转化为车辆制动性能预测的神经网络模型,再进一步使用整车部分试验结果对已得到的神经网络进行训练,得到最终的预测模型结果表明使用神经网络模型可以提高车辆制动性能预测的精度

The Application of BP Neural Network in the Calculation of the Vehicle Braking Performance

The computer simulation on JS2310 farm transport vehicle is made by setting up a dynamic model of a vehicle under braking. A comparison of the simulating result with the testing result shows that the dynamic model of vehicle under braking is correct. To improve the simulating precision of the dynamic model, artificial neural network based on back propagation algorithm(BP) is introduced.By learning form simulating results under different conditions,a BP neural network primary model is obtained.Then,some testing datas are used to train the BP neural network primary model,and the BP neural network model of the vehicle braking performance is established.The calculated precision of the BP neural network model is compared with that of the dynamic model.The result shows that the calculated precision of the BP neural network model is higher than that of the dynamic model.