基于T-S模糊故障树和贝叶斯网络的重特大交通事故成因分析

郑来; 顾鹏; 卢健

doi:10.3963/j.jssn.1674-4861.2021.04.006

基于T-S模糊故障树和贝叶斯网络的重特大交通事故成因分析

doi: 10.3963/j.jssn.1674-4861.2021.04.006

郑来^1, ,,
顾鹏¹,
卢健²

1.
哈尔滨工业大学交通科学与工程学院哈尔滨 150090
2.
公安部交通管理科学研究所江苏无锡 214151

基金项目:

国家自然科学基金面上项目 52072097

国家自然科学基金青年基金项目 71701055

详细信息

通讯作者:
郑来(1985—)，博士，副教授. 研究方向：道路交通安全理论与技术. E-mail: zhenglai@hit.edu.cn

中图分类号: U491.31
计量
- 文章访问数: 1503
- HTML全文浏览量: 524
- PDF下载量: 205
- 被引次数: 0
出版历程

A Cause Analysis of Extraordinarily Severe Traffic Crashes Based on T-S Fuzzy Fault Tree and Bayesian Network

ZHENG Lai^{1
, ,},
GU Peng¹,
LU Jian²

1.
School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin 150090, China
2.
Traffic Management Research Institute of the Ministry of Public Security, Wuxi 214151, Jiangsu, China

摘要

摘要: 重特大交通事故是最严重的交通事故类型, 为了识别此类事故的主要致因, 融合T-S模糊故障树和贝叶斯网络对其进行深入分析。建立了以重特大交通事故为顶事件, 人、车、路、环境4个因素为中间事件, 24个子因素为基本事件的T-S模糊故障树, 将其转化为贝叶斯网络, 进而双向推理基本事件的重要度和后验概率, 确定主要致因。结果表明: 融合T-S模糊故障树与贝叶斯网络的方法通过正、反向推理提高了重特大交通事故成因分析结果的准确性和可靠性, 确定了操作不当、超速、防护设施不完善、弯坡组合、路面湿滑、未按规定行驶为重特大交通事故的6个主要致因, 并对这6个主要致因之间的组合关系进一步分析, 得到了操作不当和超速对于重特大交通事故更为关键。
- 交通安全 /
- 重特大交通事故 /
- 事故成因分析 /
- T-S模糊故障树 /
- 贝叶斯网络
Abstract: The T-S fuzzy fault tree and Bayesian network are integrated for an in-depth analysis to identify the main causes of extraordinarily severe traffic crashes. A T-S fuzzy fault tree is established, with the extraordinarily severe traffic crash taken as the top event, the human, vehicle, road, and environmental factors taken as the intermediate events, and 24 sub-factors taken as the basic events. The fuzzy fault tree is transformed into a Bayesian network, and the importance and posterior probability of the basic events can be inferred biaxially to determine the main causes. The results show that the method of fusing T-S fuzzy fault tree and Bayesian network can improve the accuracy and reliability of the analysis results of the causes of extraordinarily severe traffic crashes through forward and reverse reasoning and can determine improper operation, speeding, imperfect protection facilities, and bending. Slope combination, slippery road surface, and failure to drive following regulations are the six major causes of extraordinarily severe traffic crashes. The six major causes are analyzed, revealing that improper operation and speeding are more critical for extraordinarily severe traffic crashes.
- traffic safety /
- extraordinarily severe traffic crashes /
- cause analysis of crashes /
- T-S fuzzy fault tree /
- Bayesian network

HTML全文

图 1 T-S模糊故障树与贝叶斯网络

Figure 1. T-S fuzzy fault tree and Bayesian network

下载: 全尺寸图片幻灯片

图 2 重特大交通事故成因分析T-S模糊故障树

Figure 2. T-S fuzzy fault tree for analyzing the causes of extraordinarily severe crashes

下载: 全尺寸图片幻灯片

图 3 重特大交通更事故成因分析贝叶斯网络

Figure 3. Bayesian network for analyzing the cause of extraordinarily severe crashes

下载: 全尺寸图片幻灯片

图 4 6个主要致因中不同数量的致因组合

Figure 4. Different number of causative combination of the six major causes

下载: 全尺寸图片幻灯片

图 5 重特大交通事故肇事车辆类型

Figure 5. Vehicle types of extraordinarily severe crashes

下载: 全尺寸图片幻灯片

表 1 重特大交通事故影响因素数据汇总

Table 1. Summary of factors influencing extraordinarily severe traffic crashes

因素分类/编码		子因素/编码	频数
人的因素/y₁	不良状态/y₅	疲劳驾驶/x₁	13
		酒、毒驾/x₂	5
		无证驾驶或与准驾车型不符/x₃	29
		注意力不集中/x₄	16
	不良行为/y₆	超载/x₅	99
		超速/x₆	92
		操作不当/x₇	94
		未按规定行驶/x₈	77
车辆因素/y₂	机械故障/y₇	制动不良/x₉	42
	机械故障/y₇	转向失控/x₁₀	20
	其他	机件质量问题/x₁₁	28
	其他	非法改装/x₁₂	12
道路因素/y₃	不良线形/y₈	平面转弯路段/x₁₄	13
		坡道路段/x₁₅	22
		弯坡组合路段/x₁₅	51
	路面不良	路面湿滑/x₁₆	50
	状态/y₉	施工路段/x₁₇	16
	防护设施	防护设施不完善/x₁₈	51
	隐患/y₁₀	标志标线不完善/x₁₉	63
环境因素/y₄	不良天气/y₁₁	雨/x₂₀	28
		雪/x₂₁	8
		雾/x₂₂	13
	其他	夜间无照明/x₂₃	17
	其他	视线障碍/x₂₄	8

下载: 导出CSV

表 2 根节点状态概率模糊子集

Table 2. Fuzzy subset of the state probability of root nodes

节点	$\tilde{P}\left(x_{i}=0\right)$	$\tilde{P}\left(x_{i}=0.5\right)$	$\tilde{P}\left(x_{i}=1\right)$
x₁	(0.958, 0.950, 0.943)		(0.043, 0.050, 0.058)
x₂	(0.984, 0.981, 0.978)		(0.016, 0.019, 0.022)
x₃	(0.907, 0.890, 0.874)		(0.094, 0.110, 0.127)
x₄	(0.947, 0.938, 0.929)		(0.053, 0.062, 0.071)
x₅	(0.674, 0.616, 0.558)	(0.188, 0.221, 0.254)	(0.139, 0.163, 0.187)
x₆	(0.697, 0.643, 0.589)		(0.303, 0.357, 0.411)
x₇	(0.691, 0.636, 0.581)		(0.043, 0.050, 0.058)
x₈	(0.747, 0.702, 0.657)		(0.016, 0.019, 0.022)
x₉	(0.861, 0.837, 0.813)	(0.056, 0.066, 0.076)	(0.082, 0.097, 0.112)
x₁₀	(0.934, 0.922, 0.910)		(0.066, 0.078, 0.090)
x₁₁	(0.907, 0.891, 0.875)		(0.093, 0.109, 0.125)
x₁₂	(0.960, 0.953, 0.946)		(0.040, 0.047, 0.054)
x₁₃	(0.927, 0.914, 0.901)		(0.073, 0.086, 0.099)
x₁₄	(0.877, 0.855, 0.833)		(0.123, 0.145, 0.167)
x₁₅	(0.714, 0.664, 0.614)		(0.082, 0.097, 0.112)
x₁₆	(0.720, 0.671, 0.622)		(0.066, 0.078, 0.090)
x₁₇	(0.911, 0.895, 0.879)		(0.089, 0.105, 0.121)
x₁₈	(0.714, 0.664, 0.614)		(0.286, 0.336, 0.386)
x₁₉	(0.647, 0.585, 0.523)	(0.201, 0.237, 0.273)	(0.151, 0.178, 0.205)
x₂₀	(0.844, 0.816, 0.788)	(0.095, 0.112, 0.129)	(0.061, 0.072, 0.083)
x₂₁	(0.956, 0.948, 0.940)	(0.022, 0.026, 0.030)	(0.022, 0.026, 0.030)
x₂₂	(0.928, 0.915, 0.902)	(0.022, 0.026, 0.030)	(0.050, 0.059, 0.068)
x₂₃	(0.905, 0.888, 0.871)		(0.089, 0.105, 0.121)
x₂₄	(0.955, 0.947, 0.939)		(0.286, 0.336, 0.386)

下载: 导出CSV

表 3 基本事件的关键重要度和模糊重要度

Table 3. Key importance and fuzzy importance of the basic event

基本事件	关键重要度	模糊重要度	后验概率
x1	0.000 49	0.009 67	0.050 56
x₂	0.000 18	0.009 37	0.01921
x₃	0.001 15	0.010 30	0.111 22
x₄	0.000 61	0.009 79	0.062 69
x₅	0.002 10	0.008 58	0.19367
x₆	0.005 09	0.014 10	0.360 97
x₇	0.005 23	0.014 21	0.368 05
x₈	0.003 89	0.012 92	0.301 31
x₉	0.000 78	0.008 26	0.082 29
x₁₀	0.000 79	0.010 06	0.078 87
x₁₁	0.001 15	0.010 40	0.110 21
x₁₂	0.000 46	0.009 73	0.047 52
x₁₃	0.000 86	0.009 92	0.086 96
x₁₄	0.001 56	0.010 62	0.146 61
x₁₅	0.004 63	0.013 62	0.339 73
x₁₆	0.004 49	0.013 51	0.332 66
x₁₇	0.001 08	0.010 16	0.106 17
x₁₈	0.004 64	0.013 65	0.339 74
x₁₉	0.002 38	0.008 69	0.209 29
x₂₀	0.000 81	0.008 19	0.092 82
x₂₁	0.000 21	0.007 80	0.026 24
x₂₂	0.000 38	0.007 95	0.042 93
x₂₃	0.001 18	0.010 44	0.113 25
x₂₄	0.000 52	0.009 79	0.053 59

下载: 导出CSV

表 4 基本事件的关键重要度、模糊重要度、后验概率排序

Table 4. Critical importance, fuzzy importance, and posterior probability ranking of basic events

排序	关键重要度	模糊重要度	后验概率
1	x₇	x₇	x₇
2	x₆	x₆	x₆
3	x₁₈	x₁₈	x₁₈
4	x₁₅	x₁₅	x₁₅
5	x₁₆	x₁₆	x₁₆
6	x₈	x₈	x₈
7	x₁₉	x₁₄	x₁₉
8	x₅	x₂₃	x₅
9	x₁₄	x₁₁	x₁₄
10	x₂₃	x₃	x₂₃
11	x₁₁	x₁₇	x₃
12	x₃	x₁₀	x₁₁
13	x₁₇	x₁₃	x₁₇
14	x₁₃	x₄	x₂₀
15	x₂₀	x₂₄	x₁₃
16	x₁₀	x₁₂	x₉
17	x₉	x₁	x₁₀
18	x₄	x₂	x₄
19	x₂₄	x₁₉	x₂₄
20	x₁	x₅	x₁
21	x₁₂	x₉	x₁₂
22	x₂₂	x₂₀	x₂₂
23	x₂₁	x₂₂	x₂₁
24	x₂	x₂₁	x₂

下载: 导出CSV

参考文献(20)

[1]	孟祥海. 道路交通安全技术与实践案例[M]. 北京: 人民交通出版社, 2017. MENG Xianghai. Road traffic safety technology and practice cases[M]. Beijing: People's Communications Press, 2017. (in Chinese)
[2]	许得杰, 钱勇生, 王敏, 等. 甘肃省重特大道路交通事故特征分析[J]. 中国公共安全(学术版), 2011(4): 100-103. doi: 10.3969/j.issn.1672-2396.2011.04.026 XU Dejie, QIAN Yongsheng, WANG Min, et al. Analysis of characteristics of major road traffic accidents in Gansu province[J]. China Public Safety (Academic Edition), 2011(4): 100-103. (in Chinese) doi: 10.3969/j.issn.1672-2396.2011.04.026
[3]	袁泉, 李一兵, 陈康. 引发重大交通事故的显著因素特点分析及安全对策[J]. 中国司法鉴定, 2015(5): 34-40. doi: 10.3969/j.issn.1671-2072.2015.05.006 YUAN Quan, LI Yibing, CHEN Kang. Analysis of the characteristics of the significant factors leading to major traffic accidents and safety countermeasures[J]. China Forensic Expertise, 2015(5): 34-40. (in Chinese) doi: 10.3969/j.issn.1671-2072.2015.05.006
[4]	XU Chengcheng, BAO Jie, WANG Chen, et al. Association rule analysis of factors contributing to extraordinarily severe traffic crashes in China[J]. Journal of Safety Research, 2018(67): 65-75. http://www.onacademic.com/detail/journal_1000040875882610_2077.html
[5]	ZENG Qiang, GU Weihua, ZHANG Xuan, et al. Analyzing freeway crash severity using a Bayesian spatial generalized ordered logit model with conditional autoregressive priors[J]. Accident Analysis & Prevention, 2019(127): 87-95. http://www.onacademic.com/detail/journal_1000041580367699_c5b4.html
[6]	XU Chengcheng, BAO Jie, LIU Pan, et al. Investigation of contributing factors to extremely severe traffic crashes using survival theory[J]. International Journal of Injury Control and Safety Promotion, 2018, 25(2): 141-153. doi: 10.1080/17457300.2017.1363784
[7]	由冰玉, 廉福绵, 孟祥海. 基于故障树贝叶斯网的山区高速公路事故成因分析[J]. 交通信息与安全, 2019, 37(4): 44-51. doi: 10.3963/j.issn.1674-4861.2019.04.006 YOU Bingyu, LIAN Fumian, MENG Xianghai. Cause analysis of mountain expressway accidents based on fault tree Bayesian networks[J]. Journal of Transport Information and Safety, 2019, 37(4): 44-51. (in Chinese) doi: 10.3963/j.issn.1674-4861.2019.04.006
[8]	柳昕汝. 山区高速公路伤亡事故故障树及贝叶斯网络模型[D]. 哈尔滨: 哈尔滨工业大学, 2019. LIU Xinru. Fault tree and Bayesian network model of casualty accidents on mountain expressway[D]. Harbin: Harbin Institute of Technology, 2019. (in Chinese)
[9]	杜秀丽. 重特大道路交通事故致因研究[D]. 上海: 华东师范大学, 2016. DU Xiuli. Research on the causes of major road traffic accidents[D]. Shanghai: East China Normal University, 2016. (in Chinese)
[10]	刘勇, 罗德林, 石翠, 等. 基于T-S模糊故障树的多态导航系统性能可靠性[J]. 北京航空航天大学学报, 2021, 47(2): 240-246. https://www.cnki.com.cn/Article/CJFDTOTAL-BJHK202102007.htm LIU Yong, LUO Delin, SHI Cui, et al. Performance reliability of multi-state navigation system based on T-S fuzzy fault tree[J]. Journal of Beijing University of Aeronautics and Astronautics, 2021, 47(2): 240-246. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-BJHK202102007.htm
[11]	罗彦斌, 陈建勋, 王梦恕. 基于T-S模糊故障树理论的公路隧道冻害分析方法[J]. 北京交通大学学报, 2012, 36(4): 55-60. doi: 10.3969/j.issn.1673-0291.2012.04.011 LUO Yanbin, CHEN Jianxun, WANG Mengshu. Highway tunnel freezing damage analysis method based on T-S fuzzy fault tree theory[J]. Journal of Beijing Jiaotong University, 2012, 36(4): 55-60. (in Chinese) doi: 10.3969/j.issn.1673-0291.2012.04.011
[12]	付大伟. 基于贝叶斯网络汽车起重机液压系统可靠性分析[D]. 太原: 太原科技大学, 2016. FU Dawei. Reliability analysis of hydraulic system of truck crane based on Bayesian network[D]. Taiyuan: Taiyuan University of Science and Technology, 2016. (in Chinese)
[13]	KHAKZAD N, KHAN F, AMYOTTE P. Safety analysis in process facilites: Comparison of fault tree and Bayesian network approaches[J]. Reliability Engineering and System Safety, 2011, 96(8): 925-932. doi: 10.1016/j.ress.2011.03.012
[14]	WILSON A G, HUZURBAZAR A V. Bayesian networks for multilevel system reliability[J]. Reliability Engineering and System Safety, 2007, 92(10): 1413-1420. doi: 10.1016/j.ress.2006.09.003
[15]	周忠宝, 马超群, 周经伦. 贝叶斯网络在多态系统可靠性分析中的应用[J]. 哈尔滨工业大学学报, 2009, 41(6): 232-235. doi: 10.3321/j.issn:0367-6234.2009.06.054 ZHOU Zhongbao, MA Chaoqun, ZHOU Jinglun. The application of Bayesian networks in the reliability analysis of polymorphic systems[J]. Journal of Harbin Institute of Technology, 2009, 41(6): 232-235. (in Chinese) doi: 10.3321/j.issn:0367-6234.2009.06.054
[16]	张荧驿. 基于T-S重要度和贝叶斯网络的多态液压系统可靠性分析[D]. 秦皇岛: 燕山大学, 2011. ZHANG Yingyi. Reliability analysis of multi-state hydraulic system based on T-S importance and Bayesian network[D]. Qinhuangdao: Yanshan University, 2011. (in Chinese)
[17]	陈东宁, 姚成玉, 党振. 基于T-S模糊故障树和贝叶斯网络的多态液压系统可靠性分析[J]. 中国机械工程, 2013, 24(7): 899-905. doi: 10.3969/j.issn.1004-132X.2013.07.011 CHEN Dongning, YAO Chengyu, DANG Zhen. Reliability analysis of multi-state hydraulic system based on T-S fuzzy fault tree and Bayesian network[J]. China Mechanical Engineering, 2013, 24(7): 899-905. (in Chinese) doi: 10.3969/j.issn.1004-132X.2013.07.011
[18]	陈舞, 王浩, 张国华, 等. 基于T-S模糊故障树和贝叶斯网络的隧道坍塌易发性评价[J]. 上海交通大学学报, 2020, 54(8): 820-830. https://www.cnki.com.cn/Article/CJFDTOTAL-SHJT202008009.htm CHEN Wu, WANG Hao, ZHANG Guohua, et al. Evaluation of tunnel collapse susceptibility based on T-S fuzzy fault tree and Bayesian network[J]. Journal of Shanghai Jiaotong University, 2020, 54(8): 820-830. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-SHJT202008009.htm
[19]	SUN Honghua, CHEN Hongxia, LI Chunwei. Reliability modeling and analysis of complex multi-state systems based on weighted triangular fuzzy numbers T-S fault tree[J]. International Journal of Performability Engineering, 2019, 15(6): 1662-1671. doi: 10.23940/ijpe.19.06.p17.16621671
[20]	《中国公路》编辑部. 我国道路交通事故主要成因和特点分析[J]. 中国公路, 2018(6): 26-27. https://www.cnki.com.cn/Article/CJFDTOTAL-GLZG201806017.htm China Highway Editorial Department. Analysis of the main causes and characteristics of road traffic accidents in my country[J]. China Highway, 2018(6): 26-27. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GLZG201806017.htm