基于组合模型的民航事故征候预测研究  

Prediction of Civil Aviation Accident Based on Combination Model

作　　者：吴雨桐 方学东[1] WU Yutong;FANG Xuedong(Civil Aviation Flight University of China,Guanghan 618307,China)

机构地区：[1]中国民用航空飞行学院,四川广汉618307

出　　处：《综合运输》2022年第12期112-116,共5页China Transportation Review

基　　金：国家交通战备办公室:民航机场保障能力评估研究项目(1021091);金义国际机场义亭预选场址净空保护区管控规划(H2021-46)。

摘　　要：为提高民航事故征候的预测精度,通过引入残差修正序列的灰色模型和三次指数平滑模型的预测结果,构建基于IOWA算子的有序加权组合预测模型。以2007年-2019年全国民航事故征候历史数据为实例样本,分别使用组合预测模型与单一预测模型进行拟合和预测验证。结果显示,改进灰色模型、三次指数平滑模型和组合模型的平均拟合精度分别为0.944、0.903、0.969,结合其他误差评判标准,表明组合预测模型的预测精度更高,误差更小,能有效地反映民航事故征候的发展趋势。In order to improve the prediction accuracy of the civil aviation accident,the ordered weighted combination prediction model based on the Iowa operator is constructed by introducing the grey model of the residual correction sequence and the cubic exponential smoothing method.Taking the historical data of national civil aviation accident symptoms from 2007 to 2019 as an example sample,the combined prediction model and single prediction model are used for fitting and prediction verification respectively.The results show that the average fitting accuracy of improved grey model,cubic exponential smoothing model and combined model are 0.944,0.903 and 0.969 respectively.The prediction accuracy of the combined prediction model is higher,the error is smaller,and the development trend of the civil aviation incident can be effectively reflected.

关 键 词：民航事故征候 IOWA算子 组合预测 改进灰色模型 三次指数平滑预测 

分 类 号：V328.2[航空宇航科学与技术—人机与环境工程]