用改进的前向神经网络预测铁路货运量  

An Improved Feedforward Neural Network Method for Predicting Railway Freight Transportation

作　　者：赵继阳[1] 熊天文[1] 

机构地区：[1]西南交通大学运输工程系

出　　处：《西南交通大学学报》1998年第3期254-256,共3页Journal of Southwest Jiaotong University

摘　　要：对影响铁路货运量的因素进行了分析。根据影响铁路货运量的诸因素的特点，介绍了一种改进的前向神经网络预测方法，并建立了铁路货运量前向神经网络预测模型。算例表明，其预测精度高于常规预测方法。The factors that tinfluence railway freight transportation are analyzed. An improved feedforward neural network method is introduced according to the characteristics of the factors. Finally, a model is proposed by empoying the improved feedforward neural network method for predicting railway freight transportation. An example shows that the new method can obtain higher predicting accuracy than MA and BP algorithms.

关 键 词：货物运输 神经网络 预测 铁路运输 

分 类 号：U294.13[交通运输工程—交通运输规划与管理]