基于时间序列数据挖掘的我国民航运输量预测分析  

Forecasting Analysis of Airline Passenger Volume in China Based on Time-sequence Data Mining

作　　者：刘博[1] 赵璐[1] 单曲轶[1] Liu Bo;Zhao Lu;Shan Quyi(College of Air Traffic Management of Civil Aviation University of China,Tianjin 300300 China)

机构地区：[1]中国民航大学空中交通管理学院

出　　处：《中国民航飞行学院学报》2019年第5期46-50,共5页Journal of Civil Aviation Flight University of China

摘　　要：为了精准预测我国民航运输量,基于1985-2017年我国民航运输量随机时间序列数据,运用自回归单整移动平均模型(ARIMA),借助Eviews8.0统计软件,建立预测模型。利用静态预测功能,实现了2015-2019年的点预测,整体拟合度良好,相对误差小,相对误差位于0.1%-0.9%之间,证明该模型是可行的。结果表明,ARIMA模型能够短期准确预测民航运输量,为航空公司、机场、空管等部门可持续发展提供决策依据。To accurately predict airline passenger volume in China, based on random timesequence data of airline passenger volume in China from 1985 to 2016 year, ARIMA is applied to build up prediction model, combined with the Eviews8.0 statistics software. Point prediction of 2015-2019 year is obtained with Static prediction, the overall fitting is pretty outstanding with relative error rather small with the range from 0.1%-0.9%, and it is proved that the model is feasible. The results indicated that the ARIMA model can accurately predict airline passenger volume in the short term, provide decision-making basis for the sustainable development of airlines, airports and air traffic control department.

关 键 词：民航运输量 时间序列数据 静态预测 

分 类 号：V35[航空宇航科学与技术—人机与环境工程]