基于ISM-BN的内河船舶航行风险因素研究*

doi:10.16265/j.cnki.issn1003-3033.2022.08.2009

摘要/Abstract

摘要：

为保障内河船舶航行安全,手动收集近5年长江段海事事故报告,提取代表风险因素频率的主要数据特征;结合事故调查报告分析风险要素,确定船舶航行风险评价指标体系;采用解释结构模型(ISM)构建风险因素之间的关联性模型,获得因素之间的层级关系以判定风险因素间的相关性。采用数据驱动的贝叶斯网络(BN),研究各因素对海事安全的影响,并通过敏感性分析和以往事故记录进行模型验证。结果表明:事故类型涉及到的关键风险因素有水域位置、船型、操纵执行和航道条件;通过所提方法能够识别出不同事故类型的关键风险因素,且风险模型的平均预测精度为82.87%。

关键词: 解释结构模型(ISM), 贝叶斯网络(BN), 内河船舶, 航行风险因素, 事故类型

Abstract:

To ensure the navigation safety of inland waterway vessels, the maritime accident reports of the Yangtze River in recent 5 years were collected manually, and the main data representing the frequency of risk factors were extracted. Combined with the accident investigation report, the risk factors are analyzed, and the evaluation index system of ship navigation risk is determined. The correlation modeling between risk factors was constructed by the ISM, and the hierarchical relationship between factors was obtained to determine the correlation between risk factors. The influence of various factors on maritime safety was studied by data-driven BN, and the model was verified by sensitivity analysis and previous accident records. The results show that the key risk factors involved in accident types include water location, ship type, maneuvering execution and channel conditions. The method proposed in this paper can identify the main risk factors of different accident types, and the average prediction accuracy of the risk model is 82.87%.

Key words: interpretive structural model(ISM), Bayesian network(BN), inland waterway vessel, navigational risk factor, accident type

赵建伟, 谢磊, 杨洋, 胡昕源, 欧昌奎, 曾荣. 基于ISM-BN的内河船舶航行风险因素研究*[J]. 中国安全科学学报, 2022, 32(8): 37-44.

ZHAO Jianwei, XIE Lei, YANG Yang, HU Xinyuan, OU Changkui, ZENG Rong. An ISM-BN based research on navigation risk factors of inland waterway vessels[J]. China Safety Science Journal, 2022, 32(8): 37-44.

图/表 15

图1

图2

图3

图4

图5

表1

图6

图7

表2

表3

续表3

表4

表5

表6

图8

参考文献 11

[1]	ZHANG Liye, WANG Hua, MENG Qiang, et al. Ship accident consequences and contributing factors analyses using ship accident investigation reports[J]. Proceedings of the Institution of Mechanical Engineers, Part O: Journal of Risk and Reliability, 2019, 233(1):35-47.
[2]	ROLF J, ASBJORN L. Maritime navigation accidents and risk indicators: an exploratory statistical analysis using AIS data and accident reports[J]. Reliability Engineering & System Safety, 2018, 176:174-186. doi: 10.1016/j.ress.2018.03.033
[3]	韩天园, 田顺, 吕凯光, 等. 基于文本挖掘的重特大交通事故成因网络分析[J]. 中国安全科学学报, 2021, 31(9):1-7.
	HAN Tianyuan, TIAN Shun, LYU Kaiguang, et al. Network analysis on causes for serious traffic accidents based on text mining[J]. China Safety Science Journal, 2021, 31(9):150-156. doi: 10.16265/j.cnki.issn1003-3033.2021.09.021
[4]	吴伋, 江福才, 姚厚杰, 等. 基于文本挖掘的内河船舶碰撞事故致因因素分析与风险预测[J]. 交通信息与安全, 2018, 36(3):8-18.
	WU Ji, JIANG Fucai, YAO Houjie, et al. An analysis and risk forecasting of inland ship collision based on text mining[J]. Journal of Transport Information and Safety, 2018, 36(3):8-18.
[5]	李文杰, 张浩游, 张文, 等. 基于模糊贝叶斯网络的长江上游航道整治筑物技术状况评价研究[J]. 重庆交通大学学报: 自然科学版, 2020, 39(9):112-118.
	LI Wenjie, ZHANG Haoyou, ZHANG Wen, et al. Technical status evaluation of regulation buildings in upper reaches of Yangtze River channel based on fuzzy Bayesian network[J]. Journal of Chongqing Jiaotong University:Natural Science, 2020, 39(9):112-118.
[6]	ZHANG Di, YAN Xinping, YANG Zaili, et al. Incorporation of formal safety assessment and Bayesian network in navigational risk estimation of the Yangtze River[J]. Reliability Engineering & System Safety, 2013, 118: 93-105. doi: 10.1016/j.ress.2013.04.006
[7]	FAN Shiqi, EDUARDO B, YANG Zaili, et al. Incorporation of human factors into maritime accident analysis using a data-driven Bayesian network[J]. Reliability Engineering & System Safety, 2020, 203: DOI: 10.1016/j.ress.2020.107070. doi: 10.1016/j.ress.2020.107070
[8]	李壮, 付姗姗, 高郭平, 等. 海冰影响下船舶北极水域航行风险评估方法[J]. 交通信息与安全, 2020, 38(3):92-98.
	LI Zhuang, FU Shanshan, GAO Guoping, et al. A risk assessment method of ship navigation in the arctic waters under influences of sea ice[J]. Journal of Transport Information and Safety, 2020, 38(3):92-98.
[9]	司东森, 张英俊, 郎坤. 基于改进BN的集装箱船舶碰撞事故致因分析[J]. 中国安全科学学报, 2019, 29(10):31-37. doi: 10.16265/j.cnki.issn1003-3033.2019.10.006
	SI Dongsen, ZHANG Yingjun, LANG Kun. Causation analysis of container ship collision accidents based on improved BN[J]. China Safety Science Journal, 2019, 29(10):31-37. doi: 10.16265/j.cnki.issn1003-3033.2019.10.006
[10]	徐东星, 尹勇, 张秀凤, 等. 长江干线水上交通事故的灰色分析与预测[J]. 中国航海, 2019, 42(2):59-65.
	XU Dongxing, YIN Yong, ZHANG Xiufeng, et al. Grey analysis and prediction of maritime traffic accidents on the trunk line of Yangze River[J]. Navigation of China, 2019, 42(2):59-65.
[11]	CENK S, BURAK K, ALI C et al. Analysing the effects of liquefaction on capsizing through integrating interpretive structural modelling (ISM) and fuzzy Bayesian networks (FBN)[J]. Ocean Engineering, 2020, 215: DOI: 10.1016/j.oceaneng.2020.107917. doi: 10.1016/j.oceaneng.2020.107917

变量	语言描述及相应的值
操纵执行	a代表正确		b代表有误
船员素质与能力	a代表好		b代表坏
通信交流	a代表充分		b代表不充分
安全意识	a代表好		b代表坏
船况	a代表好		b代表坏
船龄	a代表[1,10]	b代表[11,20]		c代表>20
船舶尺度	a代表小型	b代表中型		c代表大型
船型	a代表干货船	b代表散货船		c代表集装箱船
船型	d代表油船	e代表工作船		f代表其他船舶
载货影响	a代表有影响		b代表无影响
天气条件	a代表好		b代表坏
水文条件	a代表好		b代表坏
航道条件	a代表好		b代表坏
事发时间段	a代表06:00—18:00		b代表18:00—6:00
水域位置	a代表上游自然航道		b代表下游港区
水域位置	c代表下游定线制流域		d代表中游航道
船公司管理	a代表好		b代表坏
海事监管	a代表好		b代表坏
事故类型	a代表碰撞	b代表触碰		c代表自沉
事故类型	d代表搁浅	e代表火灾		f代表其他

节点	互信息	占比/%	信念方差
水域位置	0.133 46	6.060	0.007 832 3
船型	0.071 36	3.240	0.005 676 6
航道条件	0.042 41	1.920	0.005 004 1
操纵执行	0.04 002	1.810	0.004 191 3
天气条件	0.039 01	1.770	0.002 563 6
事发时间段	0.037 25	1.690	0.002 895 8
通信交流	0.036 09	1.640	0.002 687 6
船龄	0.028 11	1.280	0.000 812 2
船况	0.024 65	1.120	0.001 421 3
海事监管	0.018 96	0.861	0.001 124 9
安全意识	0.018 93	0.859	0.001 243 0
船公司管理	0.018 27	0.829	0.000 509 0
船员素质与能力	0.014 89	0.676	0.000 378 0
船舶尺度	0.012 37	0.561	0.000 234 5
水文条件	0.003 79	0.172	0.000 215 4
载货影响	0.003 55	0.161	0.000 507 8

水域位置	事故类型
水域位置	C₁	C₂	C₃	C₄	C₅	C₆
a	50.30	23.70	13.10	3.73	6.53	2.61
b	52.70	19.20	11.70	5.02	4.39	7.03
c	12.50	21.30	19.50	33.50	7.32	5.86
d	46.00	9.75	17.90	7.66	13.40	5.36

水域位置	事故类型
水域位置	C₁	C₂	C₃	C₄	C₅		C₆
船型	事故类型
船型	C₁	C₂	C₃	C₄	C₅	C₆
a	52.90	20.80	13.80	3.75	6.34	2.41
b	32.30	25.40	22.50	9.17	7.73	2.95
c	32.00	25.20	11.20	18.20	7.67	5.84
d	36.60	14.40	12.80	20.80	8.77	6.68
e	55.10	10.80	9.60	7.82	6.60	10.10
f	50.90	13.30	11.80	9.63	8.13	6.19
航道条件	事故类型
航道条件	C₁	C₂	C₃	C₄	C₅	C₆
a	52.00	18.20	11.90	6.67	6.90	4.26
b	29.00	21.80	19.60	15.40	8.09	6.01
操纵执行	事故类型
操纵执行	C₁	C₂	C₃	C₄	C₅	C₆
a	40.30	21.00	14.70	10.60	8.09	5.26
b	46.90	18.30	14.60	9.04	6.68	4.57

节点	各事故类型的R_T						均值
节点	C₁	C₂	C₃	C₄	C₅	C₆	均值
水域位置	10.47	4.89	3.20	4.06	3.04	1.31	5.41
船型	9.45	3.00	4.35	3.59	3.90	3.09	4.56
航道条件	11.50	4.80	3.85	4.37	0.60	0.88	4.33
操纵执行	12.65	5.00	0.10	0.69	0.68	3.33	3.74