河南省铁路客运量组合预测与分析  

On the combination forecasting of rail passenger numbers of Henan province

作　　者：周世国[1] 齐祥来[1] 焦成文[1] 

机构地区：[1]郑州大学数学系,河南郑州450001

出　　处：《商丘师范学院学报》2010年第9期1-6,共6页Journal of Shangqiu Normal University

基　　金：国家自然科学基金资助项目(10671183)

摘　　要：组合预测是一种综合利用各单项预测方法所提供的信息的预测模型,它是提高预测精度的有效途径.为预测河南省2005、2006两年的铁路客运量(万人),先利用直线趋势预测法,预测出河南省2005、2006两年度的铁路客运量分别为5712.285,5912.817;然后再利用灰色预测法,预测出河南省2005、2006两年度的铁路客运量分别为5888.52,6113.56;进而采用熵权法将直线趋势预测法和灰色预测法所得的预测结果,通过选取适当的权重进行加权平均,建立起组合预测模型,并预测出河南省1998-2006年的铁路客运量分别为5793.35,6005.16,而河南省2005、2006两年度的实际铁路客运量则分别为5879,6465.最后,为比较上述3种预测方法的优与劣,通过分别精确计算3种预测模型下的预测精度,表明组合预测法取得了更好的预测效果.Combination forecasting is a method which synthetically utilizes information from every single forecasting method.It is an effective way to improve the accuracy of forecasting.In order to forecast rail passenger numbers（Unit：ten thousand persons） of Henan province in the years 2005 and 2006,line tendency forecasting method and grey forecasting method are used respectively,forecasted numbers according to such two single methods are（5712.285,5912.817） and（5888.52,6113.56）;then we present a combination forecasting method,in which the coefficients of combination weights are determined by the entropy approach.by use of this method,forecasted rail passenger numbers of Henan province in the years 2005 and 2006 are（5793.35,6005.16）.However,the authentic passenger numbers in the year of 2005 and 2006 are（5879,6465）.For the sake of evaluating above mentioned three methods,we compute forecast accuracy under each method separately,the combination forecasting method is proved to achieve better results.

关 键 词：组合预测 熵 铁路客运量 

分 类 号：O212.1[理学—概率论与数理统计]