空中危险接近事件致因因素识别和作用机理分析

中国民航大学学报 ›› 2025, Vol. 43 ›› Issue (5): 16-23.

空中危险接近事件致因因素识别和作用机理分析

1.中国民航大学安全科学与工程学院，天津 300300；2.四川川航物流有限公司，成都 610000

收稿日期:2023-04-01 修回日期:2023-06-16 出版日期:2025-11-17 发布日期:2025-11-17
作者简介:陈芳（1980— ），女，湖南益阳人，教授，博士，研究方向为民航安全管理、风险管理、安全数据分析和挖掘等
基金资助:
民航安全能力建设资金项目（ASSA2021/53）

Identification of causal factors and analysis of mechanism of action of near mid-air collision incident

1. College of Safety Science and Engineering, CAUC, Tianjin 300300, China; 2. Sichuan Chuanhang Logistics Limited Company, Co.,

Ltd., Chengdu 610000, China

Received:2023-04-01 Revised:2023-06-16 Online:2025-11-17 Published:2025-11-17

摘要/Abstract

摘要：

为了帮助安全管理人员认识空中危险接近事件的发生机理，明晰致因因素之间的作用关系、重要性等，本文

提出了识别空中危险接近事件致因因素并分析其潜在作用机理的模型。利用潜在狄利克雷分配（LDA，

latent Dirichlet allocation）主题模型从事件报告和危险源清单中识别致因因素，构建决策试验和评价试验与

解释结构模型（DEMATEL-ISM，decision-making trial and evaluation laboratory and interpretative structural

model），厘清因素间的影响关系，确定各因素的重要性和属性，构建多层递阶结构模型，直观展示事件的直

接原因、间接原因和根本原因。结果表明：识别了非常态运行风险管控不力等 33 个致因因素，其中，指挥

席发布指令错误等是直接原因，直接导致事件发生；双岗制作用发挥不佳等是间接原因，直接影响较小，但

与其他因素的交互耦合作用大；安全和业务培训不到位等是根本原因，对此改善有利于从根源上降低危险

接近事件的发生。空中危险接近事件是直接、间接和根本原因共同作用的结果，基于此从空中交通管理角度

提出了具体对策建议。

关键词:

空中危险接近, 致因因素识别, 空中交通管理, 作用机理

Abstract:

In order to help safety managers understand the occurrence mechanism of near mid-air collision incident and

clarify the relationship and importance of causal factors, a model for identifying the causal factors of near mid-air

collision incident and analyzing its potential mechanism of action is proposed. Latent Dirichlet allocation (LDA)

topic model is used to identify causal factors from incident reports and lists of hazard source, the decision-making trial and evaluation laboratory and interpretative structural model (DEMATEL-ISM) is constructed, the influence relationships among factors are clarified, the importance and attributes of each factor are determined, a

multi-level hierarchical structure model is constructed to visually display the direct, indirect, and fundamental

causes of the incident. The results show that 33 causal factors are identified, including poor risk management and

control of abnormal operation, among which wrong instruction issued by command seat is the direct cause, directly leading to the occurrence of the incident. The poor performance of dual position system is an indirect cause,

which has a relatively small direct impact, but has a significant interactive coupling effect with other factors. In adequate safety and operational training are the fundamental causes, and improving them is conducive to reducing the occurrence of near mid-air collision incident from the root. Near mid-air collision incidents are the result

of the combined effects of direct, indirect, and fundamental causes, based on this, specific countermeasures and

suggestions are proposed from the perspective of air traffic management.

Key words:

near mid-air collision, causal factors identification, air traffic management, mechanism of action

中图分类号:

V267
V328

陈芳 , 李佳蕊 , 沈芮宇 .

空中危险接近事件致因因素识别和作用机理分析

[J]. 中国民航大学学报, 2025, 43(5): 16-23.

CHEN Fang , LI Jiarui , SHEN Ruiyu.

Identification of causal factors and analysis of mechanism of action of near mid-air collision incident

[J]. Journal of Civil Aviation University of China, 2025, 43(5): 16-23.

参考文献 21

[1]	ICAO. 2020 Safety Report [R]. Montreal: ICAO, 2020.
[2]	秦绪林. 相撞事故调查分析与防相撞空管策略研究[J]. 空中交通管
	理, 2009(4): 12-14.
[3]	ZHANG X G, MAHADEVAN S. Bayesian network modeling of accident
	investigation reports for aviation safety assessment[J]. Reliability Engi－
	neering and System Safety, 2021, 209: 107371.
[4]	中国民用航空局. 2019 年中国民用航空安全信息统计分析报告[R].
	北京: 中国民用航空局, 2019.
[5]	刘晓红, 杨婕. 我国双机空中相撞事故及事故征候统计分析[J].
	中国民用航空, 2012(12): 70-72.
[6]	罗晓利. 1990—2003 年中国民航 152 起小于间隔飞行事件的分类
	统计研究[J]. 中国安全科学学报, 2004, 14(12): 26-32.
[7]	陈芳, 韩美佳, 陈茜. 基于零膨胀模型的航空器危险接近事件
	安全影响因素研究[J]. 中国安全生产科学技术, 2021, 17(2): 147-151.
[8]	朱磊, 梁晓龙, 张佳强. 基于事故树和贝叶斯的军民航飞行冲突
	风险研究[J]. 火力与指挥控制, 2016, 41: 53-56.
[9]	吴奇科, 姚登凯, 甘旭升, 等. 基于 HAZOP 法的飞行冲突故障树的
	建立[J]. 火力与指挥控制, 2017, 42(2): 124-129.
[10]	WANG L L, CAO Q G, ZHOU L J. Research on the influencing factors
	in coal mine production safety based on the combination of DEMATEL
	and ISM[J]. Safety science, 2018, 103: 51-61.
[11]	陈芳, 韩美佳. 空中危险接近事件中的空管风险因素识别[J]. 安
	全与环境学报, 2021, 21(4): 1583-1591.
[12]	傅贵, 索晓, 贾清淞, 等. 10 种事故致因模型的对比研究[J]. 中
	国安全生产科学技术, 2018, 14(2): 58-63.
[13]	TANGUY L, TULECHKI N, URIELI A, et al. Natural language process－
	ing for aviation safety reports: from classification to interactive analy－
	sis[J]. Computers in Industry, 2016, 78: 80-95
[14]	王洁宁, 张聪俊, 张钰涵. 民航不安全事件报告危险源识别模型[J].
	安全与环境学报, 2020, 20(1): 186-192.
[15]	牛毅, 樊运晓, 高远. 基于数据挖掘的化工生产事故致因主题
	抽取[J]. 中国安全生产科学技术, 2019, 15(10): 165-170.
[16]	谭章禄, 王泽, 陈晓. 基于 LDA 的煤矿安全隐患主题发现研
究[J]	中国安全科学学报, 2016, 26(6): 123-128.
[17]	中国民用航空局. 民用航空器征候(征求意见稿) [EB/OL]. (2021-
	06-21)[2022-12-22]. https://www.caac.gov.cn/HDJL/YJZJ/202106/
	t20210621_208023. html.
[18]	BLEI D M, NG A Y, JORDAN M I. Latent dirichlet allocation[J]. the
	Journal of Machine Learning Research, 2003, 3: 993-1022.
[19]	SUNG Y Y, KIM S B. Topical keyphrase extraction with hierarchical
	semantic networks[J]. Decision Support Systems, 2020, 128: 113163.
[20]	YAZDI M, KHAN F, ABBASSI Ｒ, et al. Improved DEMATEL method－
	ology for effective safety management decision -making [J]. Safety
	Science, 2020, 127: 1-17.
[21]	孟超, 李晓萌, 孙玉岭, 等. 基于 ISM 和信息熵的铁路隧道塌方风
	险分析[J]. 中国安全科学学报, 2018, 28(S1): 64-70.