摘要
铁路客运量预测是铁路旅客运输生产的重要基础工作,我国铁路运输业客运量呈现出明显的季节性波动.准确预测月度铁路运量能够为我国铁路运输发展规划提供科学依据,具有非常重要的现实意义.选择了2007年1月至2014年11月的全国铁路客运量数据来建立季节时间序列SARIMA模型.经过逐期差分和季节差分后最终建立了ARIMA(1,1,2)×(1,1,1)_6模型,并选取了2008年3月至2014年11月我国铁路客运量数据的真实值和模型预测值进行了对比,平均绝对百分比误差MAPE是10.886%,模型的预测能力"优良",最后对201.5年1月-2015年3月全国铁路客运量进行了预测.
The prediction of railway passenger volume is an important base for the production of railway passenger transport.The railway transport industry of China＇s passenger volume shows significant seasonal fluctuations.Accurate prediction of monthly railway passenger volume can provide scientific basis for the development and planning of railway transportation in our country,and it has very important practical meaning.The national railway passenger volume data from January 2007 to November 2014 is chosen to establish a seasonal time series,SARIMA model.Through a period by period difference and seasonal difference,a ARIMA（1,1,2）×（1,1,1）_6 is set up.Comparing the predicted value with real value,which from the March 2008 to November 2014,the MAPE is 10.886%.The prediction ability of the model is good.In the end,the railway passenger volume from January 2015 to March 2015 is predicted.
出处
《数学的实践与认识》
北大核心
2015年第18期95-104,共10页
Mathematics in Practice and Theory
基金
教育部人文社科基金项目(13YJCZH278)
关键词
客运量预测
时间序列
SARIMA
prediction of passenger volume
time series
SARIMA
MAPE
