基于X-13A-S季节调整方法的铁路客运量预测分析

Prediction and analysis about railway passenger volume based on X-13A-S seasonal adjustment method

依据季节调整思想和X-13A-S模型理论,以传统的SARIMA模型对2000年1月—2016年7月观测值进行建模,预测2016年8—12月铁路客运量,同时用X-13A-S模型对数据中可能存在的日历效应进行季节调整,并建模预测。对比SARIMA模型,X-13A-S模型拟合效果更优,更适合我国铁路客运量的预测。之后用X-13A-S技术将所有原始数据重新建模预测,预计2017年春运后,4月份会再次出现小高峰,将比春运客流高,8月份铁路客运量达到最大。随着假期的减少,10月后客运量将下降,2018年年初迅速回升,总体呈上升趋势。

According to the seasonal adjustment and the X-13A-S model theory,firstly we model the observation data between January 2000 and July 2016 with traditional SARIMA model,and make a prediction about railway passenger volume in August 2016 to December 2016.Meanwhile,we make some seasonally adjustments for the calendar effect which possibly existed in the observation data by X13A-S model,then model and predict for them.Compared to the SARIMA model,the X-13A-S model has the better fitting results,more suitable for the forecast of Chinese railway passengers.After that,we remodel and forecast by X-13A-S technology with original observations The result of railway passenger volume is expected to have a small peak in April 2017 after Spring Festival.The transportation passengers will flow higher than the Spring Festival,and achieve a maximum of railway passenger traffic in August.With less holidays after October,railway traffic will decline,but will rebound quickly in early 2018.There is a growing trend in 2018 as a whole.