基于贝叶斯网络的电镀企业火灾事故情景推演

doi:10.16265/j.cnki.issn1003-3033.2023.02.0420

摘要/Abstract

摘要：

为有效解决电镀企业火灾事故情景演化路径复杂、火灾扑救困难等问题,提高电镀火灾应急处置能力,采用情景分析与贝叶斯网络(BN)相结合的方法,对电镀企业火灾演化路径进行推演。首先分析电镀火灾特点,确定情景状态、应急驱动、承灾要素和处置目标4种情景要素;然后基于动态BN确定情景间的关系,运用D-S证据理论处理优化,构建电镀企业火灾情景推演模型;最后将其应用于某电镀火灾实例,通过Netica软件得到各节点的状态概率,找出事故的发展趋势和情景演化路径。结果表明:该事故所有可能的演化路径中,预热槽体起火(情景S₁)是事故发生概率最大的情景,概率达81.2%;需在此处采取有效措施,使火灾消失概率达到62.3%。

关键词: 电镀企业, 火灾事故, 情景推演, 贝叶斯网络(BN), D-S证据理论

Abstract:

In order to effectively solve the problems of complex scenario evolution path and difficult firefighting in electroplating enterprises fire accident, and improve the emergency response ability of electroplating fire, the method of combining scenario analysis and BN was used to deduce the evolution path of electroplating enterprises fire. Firstly, the characteristics of electroplating fire were analyzed, and four scenario elements, including scenario status, emergency drive, disaster bearing elements and disposal objectives, were determined. Then, the relationship between scenarios was determined based on BN. The D-S evidence theory was used to process and optimize. A fire scenario deduction model was built for electroplating enterprises. Finally, taking an electroplating fire as an example, the status probability of each node was obtained through Netica software to find out the development trend and scenario evolution path of the accident. The results show that among all possible evolution paths of the accident, preheat the tank for fire (Scenario S₁) is the scenario with the highest probability of accident occurrence, with a probability of 81.2%.Effective measures should be taken here to make the probability of fire disappearance reach 62.3%.

Key words: electroplating enterprises, fire accidents, scenario deduction, Bayesian network(BN), D-S evidence theory

马港, 徐晓楠, 郭小芳, 张志珍, 徐子豪. 基于贝叶斯网络的电镀企业火灾事故情景推演[J]. 中国安全科学学报, 2023, 33(2): 202-208.

MA Gang, XU Xiaonan, GUO Xiaofang, ZHANG Zhizhen, XU Zihao. Scenario deduction of fire accidents in electroplating enterprises based on Bayesian network[J]. China Safety Science Journal, 2023, 33(2): 202-208.

图/表 12

表1

表2

表3

表4

图1

图2

图3

表5

图4

表6

表7

图5

参考文献 18

[1]	兰竹. 现代电镀技术在计算机网络工程与实践中的应用[J]. 电镀与精饰, 2021, 43(4): 55.
[2]	刘东昱, 吴奇峰. 电镀行业火灾危险性浅析[J]. 中国公共安全:综合版, 2015, 14(13): 147-149.
[3]	王德府. 电镀行业火灾原因分析和防火建议[J]. 水上消防, 2021, 41(4): 33-35.
[4]	沈建国. 电镀行业火灾特点及火灾扑救对策探讨[J]. 消防技术与产品信息, 2014, 27(5): 51-54.
	SHEN Jianguo. Fire characteristics of electroplating industry and firefighting countermeasures[J]. Fire Technique and Products Information, 2014, 27(5): 51-54.
[5]	余建荣, 王龙. 电镀企业火灾预防及扑救措施研究[J]. 化学工程与装备, 2017, 46(2): 257-258.
[6]	罗曦. 印制电路板电镀车间火灾风险分析及对策研究[D]. 广州: 华南理工大学, 2019.
	LUO Xi. Investigation on fire risk analysis and countermeasures of electroplating workshop for PCB[D]. Guangzhou: South China University of Technology, 2019.
[7]	魏长旺. 基于事故树的电镀厂房火灾危险性分析[J]. 广东化工, 2021, 48(4): 250-251.
	WEI Changwang. Fire risk analysis of electroplating factory based on fault tree[J]. Guangdong Chemical Industry, 2021, 48(4): 250-251.
[8]	李健行, 夏登友, 武旭鹏. 基于贝叶斯网络的高层建筑火灾后果预测模型[J]. 中国安全科学学报, 2013, 23(12): 54-59.
	LI Jianxing, XIA Dengyou, WU Xupeng. Bayesian Network-based model for predicting consequences of high-rise buildings fire[J]. China Safety Science Journal, 2013, 23(12): 54-59.
[9]	CHENG Hao, HADJISOPHOCLEOUS G V. The modeling of fire spread in buildings by Bayesian network[J]. Fire Safety Journal, 2009, 44(6): 901-908. doi: 10.1016/j.firesaf.2009.05.005
[10]	李健行, 夏登友, 武旭鹏. 基于知识元与动态贝叶斯网络的非常规突发灾害事故情景分析[J]. 安全与环境学报, 2014, 14(4): 165-170.
	LI Jianhang, XIA Dengyou, WU Xupeng, Scenario analysis of unconventional emergency disaster accidents based on the knowledge elements and dynamic Bayesian networks[J]. Journal of Safety and Environment, 2014, 14(4): 165-170.
[11]	KAHN H, WIENER A J. The year 2000: a framework for speculation on the next thirty-three years[M]. New York, EUA: The MacMillan Company, 1967: 20-21.
[12]	田光明. 情景分析法[J]. 晋图学刊, 2008, 24(3): 7-9,12.
	TIAN Gangming. Analysis of scenario[J]. Shanxi Library Journal, 2008, 24(3): 7-9,12.
[13]	辛晶, 周扬, 夏登友, 等. 城市商业综合体建筑火灾事故情景推演模型[J]. 灾害学, 2020, 35(3): 63-66,89.
	XIN Jing, ZHOU Yang, XIA Dengyou, et al. Scenario deduction model of urban commercial complex fire accidents[J]. Journal of Catastrophology, 2020, 35(3): 63-66,89.
[14]	屈静, 张建彬, 李旭芳, 等. 基于贝叶斯网络的输油管道泄漏事故情景推演[J]. 中国安全科学学报, 2021, 31(1): 192-198. doi: 10.16265/j.cnki.issn1003-3033.2021.01.028
	QU Jing, ZHANG Jianbin, LI Xufang, et al. Deduction of leakage accident scenarios of oil pipelines based on Bayesian network[J]. China Safety Science Journal, 2021, 31(1): 192-198. doi: 10.16265/j.cnki.issn1003-3033.2021.01.028
[15]	ANDERSEN S K. Probabilistic reasoning in intelligent systems: networks of plausible inference[J]. Artificial Intelligence, 1991, 48(1): 117-124. doi: 10.1016/0004-3702(91)90084-W
[16]	KLEIN G A, ORASANU J, CALDERWOOD R, et al. Decision making in action: models and methods[M]. Westport: Ablex Publishing, 1994: 263-264.
[17]	SHAFER G. A mathematical theory of evidence[M]. Princeton: Princeton University Press, 1976: 5-10.
[18]	王嘉玮. 基于D-S证据理论的道路交通事故判决方法研究[D]. 北京: 北京交通大学, 2021.
	WANG Jiawei. Research on traffic accident judgment method based on D-S evidence theory[D]. Beijing: Beijing Jiaotong University, 2021.

类别	要素
初期增长阶段	电加热棒起火和危险化学品起火等
充分发展阶段	沿排气管道和通风管道蔓延等
减弱阶段	火势受到控制和火灾消失等

类别	要素
建筑结构	建筑耐火等级和防火分区等
基础设施	电气线路和通风管道等
材料设备	可燃危险化学品和电加热设备等
人力资源	受困人员、救援人员

类别	要素
应急资源	水灭火系统、灭火器等
应急响应	报警时间和消防队到达现场的时间等
应急指挥	决策者应急指挥调度和应急预案等
应急救援	企业消防队和消防队员等

类别	要素
人	救援受困人员
物	隔离控制化工原料
环境	制止火灾蔓延和扑灭火灾等

情景状态S	处置目标G	承灾要素E	应急驱动R
预热槽体起火S₁	隔离可燃化工原料 G₁	化学反应材料 E₁	关闭蒸发器电源 R₁
蒸发器起火S₂	扑灭初期火灾 G₂	易燃有机物 E₂	操作人员报警 R₂
火灾消失S₃	—	—	—
火势扩大至整个铁皮棚一层S₄	制止火势扩大 G₄	楼梯通风口 E₄	内部灭火 R₄
火势纵向引燃铁皮棚二层S₅	制止火势蔓延 G₅	铁皮棚内可燃物 E₅	消防灭火系统 R₅
火灾消失S₆	—	—	—
火势突破防火分区蔓延至南一区车间S₇	救援受困人员 G₇	生产车间电气线路 E₇	消防水炮压制 R₇
火势蔓延至北一层办公区S₈	制止火势继续蔓延 G₈	办公区建筑结构 E₈	消防人员灭火 R₈
火灾消失S₉	—	—	—
火势沿电梯间向上蔓延至食堂S₁₀	制止火势继续扩大 G₁₀	石膏板分隔 E₁₀	消防水炮压制R₁₀
火灾消失S₁₁	—	—	—
火势蔓延至南二层生产线S₁₂	扑灭火灾 G₁₂	生产车间电气线路 E₁₂	消防人员灭火 R₁₂
火灾消失S₁₃	—	—	—
火灾在整个电镀厂燃烧S₁₄	扑灭火灾 G₁₄	建筑内部结构 E₁₄	消防人员灭火 R₁₄
火灾消失S₁₅	—	—	—