基于灰色模型的铁路分品类货运量预测  

Traffic Volume Forecast of Railway Different Freight Categories Based on Grey Model

作　　者：付建飞[1] 安仲文 宋小满[1] 杨瑜[1] 

机构地区：[1]中国铁道科学研究院,运输及经济研究所,北京100081 [2]中国铁路总公司,运输局,北京100081

出　　处：《交通运输工程与信息学报》2014年第3期38-42,46,共6页Journal of Transportation Engineering and Information

基　　金：中国铁道科学研究院基金(1219YJ8003)

摘　　要：铁路分品类货运量预测是制定货运营销决策的依据。作者通过基础数据处理、构建微分方程、求解模型的时间响应方程3个步骤形成了灰色预测GM(1,1)模型的算法,同时介绍了模型检验方法。基于此,依据铁路1998—2010年的实际运量数据,预测得到2011、2012年铁路分品类货运量,进而对预测结果进行检验和分析,得到模型的适用性:即灰色模型对数据序列波动较少的品类预测效果较好,而数据序列波动性大的品类预测效果一般。最后,针对部分基础数据波动较大的序列,采用指数平滑法对灰色预测的结果进行修正,提高了总体的预测效果。Different railway freight category volume forecast is a base for making freight marketing decision. A gray model algorithm was consisted of three steps： data processing,construction of differential equation, solving the time response of the model equations. The model checking method was introduced. According to the actual traffic data of Chinese railway from 1998 to 2010, 2011-2012＇s category cargo traffic volumes were predicted, and then,the prediction results were tested. The results showed that the prediction effect was better if the data sequence was less volatile. Finally, the forecast result of the grey model was amended with an exponential smoothing method for the case of fluctuation data sequence. The overall forecast effect was improved.

关 键 词：灰色模型 分品类 货运量预测 

分 类 号：U294.13[交通运输工程—交通运输规划与管理]