摘要
公路运输能力受多种因素影响,各因素的作用机制通常不能准确地用数学语言进行描述.采用广义回归神经网络(GRNN)对公路运输能力进行分析及预测.通过对2000～2008年全国公路运输能力的历史数据进行分析和处理,对网络进行训练和拟合,用2007～2008年的实际数据进行模型检验,结果表明:当光滑因子为0.1时,逼近误差为0.3%,GRNN用于公路运输能力的预测具有较好效果.
Road transport capacity is subjected to many factors,which can not be accurately described in mathematical language.GRNN was adopted to analyze and predict the freight.By analyzing and processing the historic data of road transport capacity in China during the period of 2000~2008,and by training and simulating the network,the actual data were tested with a model.The results show：When the smooth factor is 0.1,approximation error is 0.3%,GRNN has effectiveness of freight prediction.
出处
《邵阳学院学报（自然科学版）》
2010年第2期26-29,共4页
Journal of Shaoyang University：Natural Science Edition
