安全与环境工程2024，Vol.31Issue(6)：16-25,10.DOI:10.13578/j.cnki.issn.1671-1556.20230515
基于文本挖掘的冲/偏出跑道事故致因网络分析
Text mining-based causal network analysis of runway overrun/excursion accidents
摘要
Abstract
Runway excursion accidents are among the typical high-risk accident types in civil aviation operations.To effectively identify the causes of runway excursion accidents and analyze the importance and interrelationships among these causes,textual information from 827 runway/excursion accidents between 2001 and 2021 sourced from the ASN website was selected as a sample.Using text mining and social network analysis methods,a novel cause-mining approach was proposed.This approach employed the Word2Vec+TextRank combined algorithm to extract cause nodes and plot a cause network diagram.Centrality analysis and cohesive subgroup analysis methods were used to determine the key causes and the clustering relationships among the various causes.The Apriori algorithm was employed to mine strong association rules among the causes,and inadequate crew resource management was taken as an example to determine the correlation among the causes.The results indicate that there are a total of 40 causes leading to runway/excursion accidents,with wind,improper decision-making,inadequate crew resource management,and inadequate training as the key causes.Clustering relationships and strong association rules exist among the accident causes.The research findings can provide a reference for effectively preventing and controlling the occurrence of runway excursion accidents.关键词
航空安全/冲/偏出跑道事故/文本信息分析/社会网络分析/关联分析/致因分析Key words
aviation safety/runway overrun/excursion accident/text message analysis/social network analysis/correlation analysis/causal analysis分类
资源环境引用本文复制引用
刘俊杰,于佳楠,王懿..基于文本挖掘的冲/偏出跑道事故致因网络分析[J].安全与环境工程,2024,31(6):16-25,10.基金项目
中国民用航空局安全能力建设项目(ASSA2022/18、KJZ49420220016) （ASSA2022/18、KJZ49420220016）
