面向化工园区有害气体泄漏的博弈巡逻策略

doi:10.11959/j.issn.2096-6652.202338

智能科学与技术学报 ›› 2023, Vol. 5 ›› Issue (3): 366-377.doi: 10.11959/j.issn.2096-6652.202338

• 专栏：智能科技与社会计算 • 上一篇下一篇

面向化工园区有害气体泄漏的博弈巡逻策略

陈茵¹, 张立泽², 帅国华³, 陈黎丽², 王震²^,⁴

¹ 中国电信股份有限公司广西分公司，广西南宁 530022
² 杭州电子科技大学网络空间安全学院，浙江杭州 310018
³ 中国（广西）自由贸易试验区钦州港片区应急管理局，广西钦州 535008
⁴ 杭州电子科技大学卓越学院，浙江杭州 310018

修回日期:2023-05-23 出版日期:2023-09-01 发布日期:2023-09-26
作者简介:陈茵（1983- ），女，就职于中国电信股份有限公司广西分公司，主要研究方向为大数据分析、信息化项目管理
张立泽（1997- ），男，杭州电子科技大学硕士生，主要研究方向为安防巡逻
帅国华（1976- ），男，中国广西自贸区钦州港片区应急局注册安全工程师、局长，主要研究方向为化工园区安全生产监督管理
陈黎丽（1985- ），女，博士，杭州电子科技大学讲师，主要研究方向为网络安全、博弈论、数据安全
王震（1984- ），男，博士，杭州电子科技大学教授，卓越学院副院长，主要研究方向为智能博弈对抗、人工智能安全、博弈论、多智能体系统、网络科学和去中心化安全

Game patrol strategy for hazardous gas leakage in chemical parks

Yin CHEN¹, Lize ZHANG², Guohua SHUAI³, Lili CHEN², Zhen WANG²^,⁴

¹ China Telecom Corporation Limited, Guangxi Branch, Nanning 530000, China
² School of Cyberspace, Hangzhou Dianzi University, Hangzhou 311100, China
³ Qinzhou Port Emergency Bureau in China Guangxi Pilot Free Trade Zone, Qinzhou 535000, China
⁴ ZhuoYue Honors College, Hangzhou Dianzi University, Hangzhou 311100, China

Revised:2023-05-23 Online:2023-09-01 Published:2023-09-26

摘要/Abstract

摘要：

在推进智慧城市一体化的背景下，智慧化工园区的建设步伐正在逐步加快。近些年，我国化工行业事故频发，给公共安全造成了巨大的破坏与损失。因此，如何提升化工园区安全管理和应急响应能力是目前亟待解决的问题。针对该类安全问题，提出面向化工园区有害气体泄漏的博弈巡逻策略。首先，利用对流扩散模型对有害气体泄漏的过程进行描述。其次，引入博弈论对巡逻问题中攻防双方的对抗过程进行建模，将防守者对安全事故的响应时间与其收益关联。然后，提出基于多线性规划的GGC算法，对该博弈模型进行求解。最后，在3种不同规模的真实化工园区案例场景下，将该博弈模型与其他两种基础方法进行收益对比。结果表明，该模型能够有效提升防守者收益，降低攻击者收益。

关键词: 化工园区, 气体泄漏, 博弈巡逻, 博弈论, 响应时间

Abstract:

In the context of promoting the integration of smart cities, the pace of construction of smart chemical parks is gradually accelerating.Since accidents in China’s chemical industry have occurred frequently in recent years, causing great damage and loss to public safety, improving the safety management and emergency response capability of chemical parks is an urgent need to be solved.For this kind of safety problems, this paper proposed a game patrol strategy for hazardous gas leakage in chemical parks.Firstly, the convective diffusion model was used to describe the process of hazardous gas leakage.Secondly, game theory was introduced to model the confrontation process between the attacking and defending parties in the patrol problem, and the response time of the defenders to a safety incident was correlated with its benefit.Then, a multilinear programming-based GGC algorithm was proposed to solve this game model.Finally, the gains of the game model in this paper were compared with the other two basic methods in the scenarios of three real chemical park case with different sizes.The results show that the model can effectively improve the gains of the defender and reduce the gains of the attacker.

Key words: chemical park, gas leakage, game patrol, game theory, response time

中图分类号:

TP309.2

陈茵, 张立泽, 帅国华, 等. 面向化工园区有害气体泄漏的博弈巡逻策略[J]. 智能科学与技术学报, 2023, 5(3): 366-377.

Yin CHEN, Lize ZHANG, Guohua SHUAI, et al. Game patrol strategy for hazardous gas leakage in chemical parks[J]. Chinese Journal of Intelligent Science and Technology, 2023, 5(3): 366-377.

图/表 22

图1

表1

博弈收益矩阵示例"

策略	$s_{a}^{1}$	$s_{a}^{2}$
$s_{d_{1}}$	$(R_{d}^{1} \cdot M_{s_{d_{1}}, s_{a}^{1}} - L_{d}^{1} \cdot N_{s_{d_{1}}, s_{a}^{1}}, G_{a}^{1} \cdot N_{s_{d_{1}}, s_{a}^{1}} - P_{a}^{1} \cdot M_{s_{d_{1}}, s_{a}^{1}})$	$(R_{d}^{2} \cdot M_{s_{d_{1}}, s_{a}^{2}} - L_{d}^{2} \cdot N_{s_{d_{1}}, s_{a}^{2}}, G_{a}^{2} \cdot N_{s_{d_{1}}, s_{a}^{2}} - P_{a}^{2} \cdot M_{s_{d_{1}}, s_{a}^{2}})$
$s_{d_{2}}$	$(R_{d}^{1} \cdot M_{s_{d_{2}}, s_{a}^{1}} - L_{d}^{1} \cdot N_{s_{d_{2}}, s_{a}^{1}}, G_{a}^{1} \cdot N_{s_{d_{2}}, s_{a}^{1}} - P_{a}^{1} \cdot M_{s_{d_{2}}, s_{a}^{1}})$	$(R_{d}^{2} \cdot M_{s_{d_{2}}, s_{a}^{2}} - L_{d}^{2} \cdot N_{s_{d_{2}}, s_{a}^{2}}, G_{a}^{2} \cdot N_{s_{d_{2}}, s_{a}^{2}} - P_{a}^{2} \cdot M_{s_{d_{2}}, s_{a}^{2}})$

表1

图2

表2

图3

表3

钦州化工园区案例场景的收益矩阵（单位：千元）"

化工厂	$R_{d}^{l}$	$L_{d}^{l}$	$G_{a}^{l}$	$P_{a}^{l}$
A	2	15	10	4
B	2	10	6.5	4
C	2	13	8	4
D	2	12	7	4
E	2	16	9	4

表3

图4

表4

表5

图5

图6

表6

上海化工园区案例场景的收益矩阵（单位：千元）"

化工厂	$R_{d}^{l}$	$L_{d}^{l}$	$G_{a}^{l}$	$P_{a}^{l}$
A	2	10	10	4
B	2	14	12	4
C	2	11	8	4
D	2	9	6	4

表6

图7

表7

表8

图8

图9

表9

Antwerp化工园区案例场景的收益矩阵（单位：千元）"

化工厂	$R_{d}^{l}$	$L_{d}^{l}$	$G_{a}^{l}$	$P_{a}^{l}$
A	3	18	15	5
B	3	15	6	5
C	3	10	10	5
D	3	8	9	5
E	3	7	8	5
F	3	12	11	5
G	3	13	10	5
H	3	7	7	5

表9

图10

表10

表11

表12

参考文献 38

[1]	赵刚 . 城市人民警察巡逻机制建设二十年回眸与展望[J]. 公安研究, 2014(2): 25-32.
	ZHAO G . Twenty years review and prospect of urban people's police patrol mechanism[J]. Policing Studies, 2014(2): 25-32.
[2]	ZHANG L B , RENIERS G , CHEN B ,et al. CCP game:a game theoretical model for improving the scheduling of chemical cluster patrolling[J]. Reliability Engineering ＆ System Safety, 2019,191:106186.
[3]	REZAZADEH A , ZHANG L B , RENIERS G ,et al. Optimal patrol scheduling of hazardous pipelines using game theory[J]. Process Safety and Environmental Protection, 2017,109: 242-256.
[4]	陈斐然, 陈彬, 朱正秋 ,等. 基于博弈论的化工园区合作巡逻策略研究[J]. 系统仿真学报, 2020,32(10): 1903-1909.
	CHEN F R , CHEN B , ZHU Z Q ,et al. Chemical cluster cooperative patrolling strategy based on game theory[J]. Journal of System Simulation, 2020,32(10): 1903-1909.
[5]	GHOSH S , VARAKANTHAM P . Dispatch guided allocation optimization for effective emergency response[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2018,32(1).
[6]	LEI C , LIN W H , MIAO L X . A stochastic emergency vehicle redeployment model for an effective response to traffic incidents[J]. IEEE Transactions on Intelligent Transportation Systems, 2014,16(2): 898-909.
[7]	LARSON R C . Urban police patrol analysis[M]. Cambridge: MIT Press, 1972.
[8]	LIBERATORE F , CAMACHO-COLLADOS M . A comparison of local search methods for the multicriteria police districting problem on graph[J]. Mathematical Problems in Engineering, 2016,2016: 1-13.
[9]	KUO P F , LORD D , WALDEN T D . Using geographical information systems to organize police patrol routes effectively by grouping hotspots of crash and crime data[J]. Journal of Transport Geography, 2013,30: 138-148.
[10]	COUPE R T , GIRLING A J . Modelling police success in catching burglars in the act[J]. Omega, 2001,29(1): 19-27.
[11]	CAMACHO-COLLADOS M , LIBERATORE F . A decision support system for predictive police patrolling[J]. Decision Support Systems, 2015,75: 25-37.
[12]	D'AMICO S J , WANG S J , BATTA R ,et al. A simulated annealing approach to police district design[J]. Computers ＆ Operations Research, 2002,29(6): 667-684.
[13]	郝建业, 邵坤, 李凯 ,等. 博弈智能的研究与应用[J]. 中国科学:信息科学,已录用,doi:10.1360/SSI-2023-0010.
	HAO J Y , SHAO K , LI K ,et al. Research and applications of game intelligence[J]. Scientia Sinica(Informationis),accepted,doi:10.1360/SSI-2023-0010.
[14]	VERGASSOLA M , VILLERMAUX E , SHRAIMAN B I . 'Infotaxis' as a strategy for searching without gradients[J]. Nature, 2007,445(7126): 406-409.
[15]	RENIERS G , DULLAERT W , KAREL S . Domino effects within a chemical cluster:a game-theoretical modeling approach by using Nash-equilibrium[J]. Journal of Hazardous Materials, 2009,167(1-3): 289-293.
[16]	RENIERS G , SOUDAN K . A game-theoretical approach for reciprocal security-related prevention investment decisions[J]. Reliability Engineering ＆ System Safety, 2010,95(1): 1-9.
[17]	RENIERS G . An external domino effects investment approach to improve cross-plant safety within chemical clusters[J]. Journal of Hazardous Materials, 2010,177(1-3): 167-174.
[18]	RENIERS G , CUYPERS S , PAVLOVA Y . A game-theory based multiplant collaboration model (MCM) for cross-plant prevention in a chemical cluster[J]. Journal of Hazardous Materials, 2012,209-210: 164-176.
[19]	ZHANG L B , RENIERS G , CHEN B ,et al. Integrating the API SRA methodology and game theory for improving chemical plant protection[J]. Journal of Loss Prevention in the Process Industries, 2018,51: 8-16.
[20]	ZHANG L B , RENIERS G . A game-theoretical model to improve process plant protection from terrorist attacks[J]. Risk Analysis, 2016,36(12): 2285-2297.
[21]	ZHU Z Q , CHEN B , RENIERS G ,et al. Playing chemical plant environmental protection games with historical monitoring data[J]. International Journal of Environmental Research and Public Health, 2017,14(10): 1155.
[22]	FENG Q L , CAI H , CHEN Z L ,et al. Using game theory to optimize allocation of defensive resources to protect multiple chemical facilities in a city against terrorist attacks[J]. Journal of Loss Prevention in the Process Industries, 2016,43: 614-628.
[23]	SAMANTA S , SEN G , GHOSH S K . A literature review on police patrolling problems[J]. Annals of Operations Research, 2022,316(2): 1063-1106.
[24]	DANTZIG G , FULKERSON R , JOHNSON S . Solution of a largescale traveling-salesman problem[J]. Journal of the Operations Research Society of America, 1954,2(4): 393-410.
[25]	LI W Q , CHEN D H , LE J J . Robot patrol path planning based on combined deep reinforcement learning[C]// Proceedings of 2018 IEEE Intl Conf on Parallel ＆ Distributed Processing with Applications,Ubiquitous Computing ＆ Communications,Big Data ＆ Cloud Computing,Social Computing ＆ Networking,Sustainable Computing ＆ Communications. Piscataway:IEEE Press, 2018: 659-666.
[26]	CAMACHO-COLLADOS M , LIBERATORE F , ANGULO J M . A multi-criteria police districting problem for the efficient and effective design of patrol sector[J]. European Journal of Operational Research, 2015,246(2): 674-684.
[27]	CHEN H F , CHENG T , YE X Y . Designing efficient and balanced police patrol districts on an urban street network[J]. International Journal of Geographical Information Science, 2019,33(2): 269-290.
[28]	MENG J X , LU X Z . The application of graph theory method in the problem of traffic patrol police service platform's site selection[J]. Applied Mechanics and Materials, 2012,246/247: 723-727.
[29]	ROSENSHINE M . Contributions to a theory of patrol scheduling[J]. Journal of the Operational Research Society, 1970,21(1): 99-106.
[30]	OLSON D G , WRIGHT G P . Models for allocating police preventive patrol effort[J]. Journal of the Operational Research Society, 1975,26(4): 703-715.
[31]	XIE Y , GU Y J . Police patrol path planning using stackelberg equilibrium based multiagent reinforcement learning[J]. Transactions of Beijing Institute of Technology, 2017,37(1): 93-99.
[32]	LIN E S , AGMON N , KRAUS S . Multi-robot adversarial patrolling:handling sequential attacks[J]. Artificial Intelligence, 2019,274: 1-25.
[33]	BUERMANN J , ZHANG J . Multi-robot adversarial patrolling strategies via lattice paths[J]. Artificial Intelligence, 2022,311:103769.
[34]	HUANG L , ZHOU M C , HAO K R ,et al. Multirobot cooperative patrolling strategy for moving objects[J]. IEEE Transactions on Systems,Man,and Cybernetics:Systems, 2022,53(5): 2995-3007.
[35]	JANA M , VACHHANI L , SINHA A . A deep reinforcement learning approach for multi-agent mobile robot patrolling[J]. International Journal of Intelligent Robotics and Applications, 2022,6(4): 724-745.
[36]	HUTCHINSON M , OH H , CHEN W H . Entrotaxis as a strategy for autonomous search and source reconstruction in turbulent conditions[J]. Information Fusion, 2018,42: 179-189.
[37]	CONITZER V , SANDHOLM T . Computing the optimal strategy to commit to[C]// Proceedings of the 7th ACM Conference on Electronic Commerce. New York:ACM, 2006: 82-90.
[38]	ZHANG Y P , XIAO Y , WANG Y ,et al. Bio-inspired patrolling scheme design in wireless and mobile sensor and robot networks[J]. Wireless Personal Communications, 2017,92: 1303-1332.

策略编号	巡逻策略	选择概率
0	Cr1-B1-A1-B1-A1	0.5
1	Cr1-B1-Cr1-B1-Cr1-B1-A1-B1	0.5

策略编号	攻击策略
0	（A1，3，3）
1	（A1，4，3）
2	（A1，5，3）
3	（A1，7，3）

策略编号	巡逻策略	选择概率
0	Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-B1-Cr1-B1	0.0163
1	Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-B1-Cr1-B1-B2	0.0599
2	Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-B1-B2-B1	0.1299
3	Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-B1-B3-B2	0.0058
4	Cr1-A1-Cr1-A1-Cr1-A1-Cr1-A1-Cr1-Cr2-B2-Cr2-B1	0.0380
5	Cr1-A1-Cr1-A1-Cr1-Cr2-B2-B1-B2	0.0271
6	Cr1-A1-Cr1-B1-B2-B1-B2	0.0459
7	Cr1-A1-Cr1-Cr2-B2-B1-B2	0.0545
8	Cr1-A1-Cr1-Cr2-C2-Cr2-B3-B2	0.0172
9	Cr1-B1-Cr1-B1-B2-B1-B2	0.0260
10	Cr1-B1-Cr2-C2-Cr2-B1	0.1086
11	Cr1-B1-Cr2-C2-Cr2-C2-Cr2-B2-Cr2	0.0630
12	Cr1-C1-Cr2-C2-Cr2-B2-Cr2-B1	0.0358
13	Cr1-Cr2-B2-B1-B2	0.1433
14	Cr1-Cr2-B3-Cr3-B3-B1	0.1433
15	Cr1-Cr2-C2-Cr2-B2-B1-Cr1	0.0146
16	Cr1-Cr2-C2-Cr2-B3-Cr3-Cr2	0.0372
17	Cr1-Cr2-Cr1-B1-Cr1	0.0334

策略编号	攻击策略
0	（B2，0，10）
1	（B3，3，10）
2	（B3，4，10）
3	（B3，5，10）

策略编号	巡逻策略	选择概率
0	Cr1-A1-A2-A1-A2-A1-A2-A1	0.2018
1	Cr1-A2-Cr1-Cr2-Cr1-A2-A1-A2-A1-A2-A1	0.2018
2	Cr1-Cr2-B1-Cr2-Cr1-A1-Cr1-Cr2-E1	0.0242
3	Cr1-Cr2-B1-Cr2-Cr1-A2-Cr1-Cr2-Cr3-Cr5	0.0242
4	Cr1-Cr2-B1-Cr2-Cr1-Cr2-Cr1-A2-Cr1-A2-A1	0.1251
5	Cr1-Cr2-Cr1-A2-Cr1-Cr2-E1-Cr2-Cr3	0.0767
6	Cr1-Cr2-Cr1-Cr2-B1-Cr2-Cr1-A1-Cr1	0.1209
7	Cr1-Cr2-Cr1-Cr2-E1-Cr2-Cr1-A1-Cr1	0.0155
8	Cr1-Cr2-Cr1-Cr2-E1-Cr2-Cr1-A2-Cr1-A2-A1	0.0913
9	Cr1-Cr2-Cr1-Cr2-Cr1-A1-Cr1-Cr2-B1-Cr3	0.0767
10	Cr1-Cr2-Cr3-B1-Cr3-B1-Cr2-Cr1-A2-Cr1	0.0417

面向化工园区有害气体泄漏的博弈巡逻策略

Game patrol strategy for hazardous gas leakage in chemical parks

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 22

参考文献 38

相关文章 1

Metrics

推荐阅读 0

化工园区	节点数	连边数	化工厂节点数	道路交叉口数
钦州化工园区	7	7	5	2
上海化工园区	10	15	7	3
Antwerp化工园区	16	25	9	7

策略编号	攻击策略
0	（A1，0，10）
1	（A1，1，10）

园区名	随机方法	基于GGP的方法	博弈方法
钦州化工园区	（-8.42，3.52）	（-6.86，2.78）	（-6.55，2.38）
上海化工园区	（-11.07，4.87）	（-9.40，3.68）	（-8.27，3.13）
Antwerp化工园区	（-13.53，6.11）	（-9.81，5.09）	（-9.34，4.95）