基于层次结构模型的RBF神经网络货运量预测方法

为了有效地进行交通货运量预测,通过对货运量影响因素的分析,建立了关于货运量影响因素的层次结构模型,并根据该模型构建基于RBF神经网络的货运量预测方法。用我国1985~2004年的货运量对该神经网络进行训练和预测,同时与BP神经网络预测方法进行比较。结果表明,该方法具有更快的运算速度和更高的精度,具有很好的预测能力和应用价值。

The Radial Basis Function Neural Network Model for Freight Volume Forecast Based on Hierarchy Configuration Model

To forecast the fright volume more effectively,the factors influencing freight volume were andalyzed and an AHP model were established.Based on this model,a Radial Basis Function neural network model for freight volume forecasting is suggested.The historical statistical data from 1985 to 2000 are used to train the RBF ANN,and then the historical data from 2001 to 2004 are used to check the RBF ANN model.Comparing with BP ANN with same set data,the RBF ANN converges more quickly and gives a more accurate result for prediction.