基于ARIMA模型的残差修正的航线运量预测方法  

Airline Transport Demand Forecast by Using ARIMA and Residual Improved Model

作　　者：景崇毅[1] 周慧艳[2] 石丽娜[3] 

机构地区：[1]中国民航飞行学院航空运输管理学院,四川广汉618307 [2]上海工程技术大学航空运输学院,上海201620 [3]上海工程技术大学飞行学院,上海201620

出　　处：《工业工程》2010年第1期74-79,共6页Industrial Engineering Journal

基　　金：国家自然科学基金资助项目(60776820);中国民航飞行学院自然科学基金资助项目(J2008-28)

摘　　要：针对民航运量数据提出对主体ARIMA模型的残差序列进行二次建模以修正原模型的精度和适应性,引入遗传进化因子来估计二次建模的参数,利用一个典型的民航运量时间序列进行建模仿真,通过对仿真建模的结果评价表明该种建模方法是有效的,能够提高模型的预测精度和改进模型的适应性。Due to the seasonal characteristics of civil aviation transport demands, a satisfactory estimation accuracy cannot be obtained by an ARIMA model only. Thus, a residual secondary modeling method is pro- posed to improve the accuracy and adaptability of the primary ARIMA model. Further, a genetic operator is introduced to the solution process of the model. With the typical data from the real world, test is conducted by simulation. Results show that the forecasting accuracy and adaptability are significantly improved. In other words, the proposed method is effective.

关 键 词：航线运量 ARIMA模型 遗传进化算子 

分 类 号：F272.1[经济管理—企业管理]