论文部分内容阅读
在分析货运量影响因素的基础上,利用BP神经网络建立新疆货运量时间序列预测网络结构模型。利用1995～2006年新疆货运量历史数据,对模型进行训练和拟合,再选用2007～2008年的历史数据作为网络模型检验样本,同时采用移动平均法、指数平滑法对新疆货运量进行预测,并对预测结果作对比分析。研究表明,采用BP神经网络预测新疆货运量比采用其他预测方法更加准确。
Based on the analysis of the influencing factors of freight volume, the time series forecasting network structure model of freight volume in Xinjiang is established by BP neural network. Using the historical data of freight volume from 1995 to 2006 in Xinjiang, the model is trained and fitted. The historical data from 2007 to 2008 are selected as the test samples for the network model. Meanwhile, the moving average method and exponential smoothing method are used to forecast the freight volume in Xinjiang, And the prediction results for comparative analysis. The research shows that using BP neural network to predict the freight volume in Xinjiang is more accurate than using other forecasting methods.