孙瑞山;陈雄;梁妍;

• 1:中国民航大学飞行分校

摘要(Abstract)：

为检测与分析飞机进近着陆阶段可能引发安全风险的运行异常现象，基于高斯混合模型提出了一种进近着陆阶段飞机运行异常检测方法。首先，将进近着陆阶段飞行数据输入期望最大算法构建高斯混合模型，研究各高斯分量随时间的变化特性，结合所输入的数据在各高斯分量概率密度函数中的计算结果，识别进近着陆阶段存在的飞行数据异常航段。然后，通过复核数据异常航段的原始飞行数据，检测与分析飞机进近着陆阶段可能引发安全风险的运行异常现象。最后，利用该方法检测与分析了462个实际运行航段中进近着陆阶段的运行异常。结果表明：当文中模型对应的BIC数值最小时，高斯分量个数为21,其中有11个高斯分量的从属度在接地后迅速衰减至0;在划定3个不同的检测阈值后，分别识别出27、15、8个飞行数据异常航段，最终通过复核检测出3个进近着陆阶段的运行异常现象。

关键词(KeyWords)： 安全社会工程;飞行安全;进近着陆阶段;高斯混合模型;异常检测

Abstract:

Keywords:

基金项目(Foundation):

作者(Authors): 孙瑞山;陈雄;梁妍;

DOI: 10.13637/j.issn.1009-6094.2021.0198

参考文献(References)：

• [1] 彭宇，何永福，王少军，等.飞行数据异常检测技术综述[J].仪器仪表学报，2019,40(3):1-13.PENG Y,HE Y F,WANG S J,et al.Flight data anomaly detection:a survey[J].Chinese Journal of Scientific Instrument,2019,40(3):1-13.
• [2] 中国民用航空局飞行标准司.飞行品质监控(FOQA)实施与管理：AC-121/135-FS-2012-45R1[S].北京：中国民航局，2015:21-25.Flight Standards Department of Civil Aviation Administration ofChina.The implementation and management of flight operational quality assurance (FOQA):AC-121/135-FS-2012-45R1[S].Beijing:Civil Aviation Administration of China,2015:21-25.
• [3] OEHLING J,BARRY D J.Using machine learning methods in airline flight data monitoring to generate new operational safety knowledge from existing data[J].Safety Science,2019,114:89-104.
• [4] MELNYK I,BANERJEE A,MATTHEWS B,et al.Semi-markov switching vector autoregressive model-based anomaly detection in aviation systems[C]//Proceedings of the 22nd ACM SIGKDD International Conference.New York:Association for Computing Machinery,2016:1065-1074.
• [5] KHALASTCHI E,KALECH M,KAMINKA G A,et al.Online data-driven anomaly detection in autonomous robots[J].Knowledge and Information Systems,2015,43(3):657-688.
• [6] 费思邈，霍琳，王亮，等.基于聚类分析的飞行数据异常检测方法[C]//2015航空试验测试技术学术交流会论文集.北京：中国航空学会，2015:277-279,302.FEI S M,HUO L,WANG L,et al.Anormaly detection method of flight data based on clustering analysis[C]//Proceedings of 2015 Academic Conference on Aviation Test Technology.Beijing:Chinese Society of Aeronautics and Astronautics,2015:277-279,302.
• [7] 何永福，王少军，王文娟，等.基于过采样投影近似基追踪的无人机异常检测[J].仪器仪表学报，2016,37(7):1468-1476.HE Y F,WANG S J,WANG W J,et al.UAV anomaly detection based on oversampling projection approximation basis pursuit[J].Chinese Journal of Scientific Instrument,2016,37(7):1468-1476.
• [8] 周德新，刘宸宇，牛亚月.关于飞行发动机数据异常实时检测仿真研究[J].计算机仿真，2016,33(9):118-122.ZHOU D X,LIU C Y,NIU Y Y.Simulation research about flight engine data anomaly detection in real time[J].Computer Simulation,2016,33(9):118-122.
• [9] 周志华.机器学习[M].北京：清华大学出版社，2016:206-211.ZHOU Z H.Machine learning[M].Beijing:Qsinghua University Press,2016:206-211.
• [10] DAEMI A,KODAMANA H,HUANG B.Gaussian process modelling with gaussian mixture likelihood[J].Journal of Process Control,2019,81:209-220.
• [11] 朱婷，赵来军，王旭磊.基于贝叶斯网络的危险化学品道路运输事故分析[J].安全与环境学报，2016,16(2):53-60.ZHU T,ZHAO L J,WANG X L.Analysis of the hazardous material transportation accidents based on the Bayesian network method[J].Journal of Safety and Environment,2016,16(2):53-60.
• [12] 王爱平，张功营，刘方.EM算法研究与应用[J].计算机技术与发展，2009,19(9):108-110.WANG A P,ZHANG G Y,LIU F.Research and application of EM algorithm[J].Computer Technology and Development,2009,19(9):108-110.
• [13] LUCKE J,FORSTER D.K-means as a vriational EM approximation of Gaussian mixture models[J].Pattern Recognition Letters,2019,125:349-356.
• [14] 刘通，付锐，张名芳，等.融合K-means与高斯混合模型的驾驶风格聚类研究[J].中国安全科学学报，2019,29(12):40-45.LIU T,FU R,ZHANG M F,et al.Study on driving style clustering based on K-means and Gaussian mixture model[J].China Safety Science Journal,2019,29(12):40-45.
• [15] The Boeing Company.B737 flight crew operations manual[M].Chicago:The Boeing Company,2009.