论文部分内容阅读
随着列车运行速度和轴重的提高,轮轨系统的磨损越来越严重,其中曲线半径、轴重和运行速度是影响轮轨磨损的重要因素。建立了钢轨磨损量影响规律的径向BP基函数神经网络模型,该网络具有3路输入,3个神经层;在JD-1大型轮轨模拟试验机上通过改变试验参数进行钢轨磨损试验,获得不同试验参数下的钢轨磨损量;以钢轨磨损数据作为BP神经网络的目标样本,对不同试验参数下的磨损量进行了预测。结果表明,模型可较准确地计算轮轨冲角和速度对钢轨磨损量的影响规律,利用BP神经网络对钢轨磨损量预测具有较高的精度,可在一定程度上验证试验结果。
With the increase of train running speed and axle load, the wear of wheel-rail system becomes more and more serious. Among them, the curve radius, axle load and running speed are the important factors that affect the wheel-rail wear. The radial BP basis function neural network model which has the influence law of the rail wear quantity is established. The network has 3 inputs and 3 neural layers. The rail wear test is carried out by changing the test parameters on the JD-1 large wheel and rail simulation testing machine, The amount of rail wear under test parameters; the rail wear data as the target sample of BP neural network, the wear amount under different test parameters was predicted. The results show that the model can accurately calculate the impact of wheel angle and angle on the rail wear. The BP neural network has a high accuracy in predicting the rail wear, which can verify the test results to a certain extent.