基于GA-PS的三维空间源项反演算法

doi:10.12382/bgxb.2022.0272

摘要/Abstract

摘要：

针对当前化学危害事件中的危害源定位及源强反算仍以二维平面为主的现状,在分析高维度源项反演寻优特性基础上,利用多参数条件下遗传算法全局寻优和模式搜索算法局部寻优性能优势,提出一种基于遗传算法-模式搜索算法的改进遗传模式搜索算法。仿真实验结果表明,该算法利用一组地面或空中探测器在三维空间中采集到的危害物质浓度数据,能够实现对危害源三维空间定位和源强反算。通过SF₆外场试验实测数据的反算求解,验证了该算法能够实现对大气环境稳定条件下的危害源空间位置、源强的有效反算,反算精度在30%以内。

关键词: 化学危害事件, 三维空间, 遗传算法, 模式搜索算法, 源项反演

Abstract:

In response to the current dominance of hazard source location and source intensity inverse calculation in chemical hazards being confined to the two-dimensional plane, this study focuses on high-dimensional source inversion optimization characteristics. To address this limitation, we propose the Modified Genetic Algorithm and Pattern Search algorithm based on Genetic Algorithm-Pattern Search Algorithm, utilizing genetic algorithms for global optimization and leveraging the pattern search algorithm’s local optimization performance. The simulation results show that the algorithm can realize the three-dimensional spatial location of the hazard source and the inverse calculation of the source intensity using the concentration data of hazardous substances from a group of ground or air detectors in the three-dimensional space. Through inverse calculation of the measured data of the SF6 field experiment, it is verified that the algorithm can realize effective inverse calculation of the spatial location and source intensity of the hazard source under stable atmospheric conditions, achieving an inverse calculation accuracy within 30%.

Key words: chemical hazard, three-dimensional space, genetic algorithm, pattern search algorithm, source inversion

中图分类号:

E91
X928.5

陈松, 朱东升, 左钦文, 韩朝帅. 基于GA-PS的三维空间源项反演算法[J]. 兵工学报, 2023, 44(8): 2503-2520.

CHEN Song, ZHU Dongsheng, ZUO Qinwen, HAN Chaoshuai. GA-PS Based Three-dimensional Space Source Inversion Algorithm[J]. Acta Armamentarii, 2023, 44(8): 2503-2520.

图/表 36

参考文献 28

[1]	郭立军, 曹霞, 宋剑. 化学事故危害后果评估与应急救援的辅助决策系统设计[J]. 中国安全生产科学技术, 2011, 7(6):33-37.
	GUO L J, CAO X, SONG J. Design of auxiliary decision-making system for chemical accident hazard consequence assessment and emergency rescue[J]. Journal of Safety Science and Technology, 2011, 7(6):33-37. (in Chinese)
[2]	王奋伟, 梁赤勇, 王男. 核生化灾害模拟与仿真技术与装备研究综述[C]//第一届中国指挥控制大会论文集. 北京: 中国指挥与控制学会, 2013:650-653.
	WANG F W, LIANG C Y, WANG N. Review on simulation technology and equipment research of nuclear, biological and chemical disaster[C]//Proceedings of the First China Command and Control Conference. Beijing: Chinese Institute of Command and Control, 2013:650-653.
[3]	韩朝帅, 诸雪征, 顾进, 等. 核化危害源项反演技术现状及研究展望[J]. 化工环保, 2021, 41(2):146-155. doi: 10.3969/j.issn.1006-1878.2021.02.004
	HAN C S, ZHU X Z, GU J, et al. Current situation and research prospect on source term inversion technology of nuclear and chemical hazards[J]. Environmental Protection of Chemical Industry, 2021, 41(2):146-155. (in Chinese) doi: 10.3969/j.issn.1006-1878.2021.02.004
[4]	李颢青. 工业有毒气体的溯源模型研究[D]. 大连: 大连理工大学, 2018.
	LI H Q. Research on the traceability model of industrial toxic gases[D]. Dalian: Dalian University of Technology, 2018. (in Chinese)
[5]	HK A, ESY B, DS A. Deep neural networks for source tracking of chemical leaks and improved chemical process safety[J]. Computer Aided Chemical Engineering, 2018, 44:2359-2364.
[6]	QIU S H, CHEN B, WANG R X, et al. Atmospheric dispersion prediction and source estimation of hazardous gas using artificial neural network, particle swarm optimization and expectation maximization[J]. Atmospheric Environment, 2018, 178:158-163. doi: 10.1016/j.atmosenv.2018.01.056 URL
[7]	YAN Y T, ZHANG R, WANG J, et al. Modified PSO algorithms with “request and reset” for leak source localization using multiple robots[J]. Neurocomputing, 2018, 292:82-90. doi: 10.1016/j.neucom.2018.02.078 URL
[8]	KONG Y G, GUAN M, ZHENG S, et al. Locating hazardous chemical leakage source based on cooperative moving and fixing sensors[J]. Sensors, 2019, 19(5):1092. doi: 10.3390/s19051092 URL
[9]	MAO S S, LANG J L, CHEN T, et al. Improving source inversion performance of airborne pollutant emissions by modifying atmospheric dispersion scheme through sensitivity analysis combined with optimization model[J]. Environmental Pollution, 2021, 284(4/5):117186. doi: 10.1016/j.envpol.2021.117186 URL
[10]	王煜薇. 基于混合遗传-Nelder Mead单纯形算法的源强及位置反算[D]. 北京: 北京化工大学, 2011.
	WANG Y W. Estimation of source position and intensity in chemical releases by a combinded Genetic-Nelder Mead simplex algorithm[D]. Beijing: Beijing University of Chemical Technology, 2011. (in Chinese)
[11]	WANG J L, LIU J J, WANG B, et al. A new method for multi-point pollution source identification[J]. Atmospheric and Oceanic Science Letters, 2021, 14(6):56-61.
[12]	史阳. 有毒气体罐车运输泄漏的源强及位置反算研究[D]. 兰州: 兰州交通大学, 2013.
	SHI Y. Research on back-calculation of source intensity and position of toxic gas dispersion from road tank car[D]. Lanzhou: Lanzhou Jiaotong University, 2013. (in Chinese)
[13]	陈增强. 混合遗传—模式搜索算法在毒气泄漏中的源强反算研究[D]. 北京: 北京化工大学, 2009.
	CHEN Z Q. Research on source inversion problem in the leakage of toxic gases based on hybrid genetic-pattern search algorithm[D]. Beijing: Beijing University of Chemical Technology, 2009. (in Chinese)
[14]	郑小平, 陈增强. 模式搜索算法在毒气泄漏中的源强反算[J]. 中国安全科学学报, 2010, 20(5):29-34.
	ZHENG X P, CHEN Z Q. Back calculation of source strength and location of toxic gases releasing based on Pattern Search method[J]. China Safety Science Journal, 2010, 20(5): 29-34. (in Chinese)
[15]	胡峰, 郎建垒, 毛书帅, 等. 典型优化目标函数下源参数反演性能对比研究[J]. 中国环境科学, 2021, 41(5):2081-2089.
	HU F, LANG J L, MAO S S, et al. Comparative study on inversion performance of source parameters under typical optimization objective function[J]. China Environmental Science, 2021, 41(5):2081-2089. (in Chinese)
[16]	BARAD M L. Project prairie grass: a field program in diffusion: AF-152572[R]. Bedford, UK: Geophysical Research Air Force Cambridge Research Laboratories, 1958.
[17]	DAI C H, CHEN W R, ZHU Y F, et al. Seeker optimization algorithm for optimal reactive power dispatch[J]. IEEE Transactions on Power Systems, 2009, 24(3):1218-1231. doi: 10.1109/TPWRS.2009.2021226 URL
[18]	郑树泉. 工业智能技术与应用[M]. 上海: 上海科学技术出版社, 2019.
	ZHENG S Q. Industrial intelligence technology and application[M]. Shanghai: Shanghai Science and Technology Press, 2019. (in Chinese)
[19]	雷英杰, 张善文. MATLAB遗传算法工具箱及应用[M]. 西安: 西安电子科技大学出版社, 2014.
	LEI Y J, ZHANG S W. Matlab genetic algorithm toolbox and application[M]. Xi’an: Xidian University Press, 2014. (in Chinese)
[20]	HOOKE R, JEEVES T A. "Direct Search" solution of numerical and statistical problems[J]. Journal of the Acm, 1961, 8(2):212-229. doi: 10.1145/321062.321069 URL
[21]	庄正宁. 环境工程基础[M]. 北京: 中国电力出版社, 2006.
	ZHUANG Z N. Fundamentals of environmental engineering[M]. Beijing: China Power Press, 2006. (in Chinese)
[22]	制定地方大气污染物排放标准的技术方法: GB/T 3840—1991[S]. 北京: 国家技术监督局/国家环境保护局, 1991.
	Technical methods for formulating local air pollutant emission standards:GB/T 3840—1991[S]. Beijing: State Bureau of Technical Supervision/State Bureau of Environmental Protection, 1991. (in Chinese)
[23]	周明, 孙树栋. 遗传算法原理及应用[M]. 北京: 国防工业出版社, 1999.
	ZHOU M, SUN S D. Genetic algorithms: theory and application[M]. Beijing: National Defense Industry Press, 1999. (in Chinese)
[24]	陈增强. 危险化学品泄漏源的定位研究[D]. 北京: 北京化工大学, 2013.
	CHEN Z Q. Research on source identification approaches for hazardous chemical releases[D]. Beijing: Beijing University of Chemical Technology, 2013. (in Chinese)
[25]	崔继宪. 突发大气污染事故快速溯源与应急评估技术研究及验证[D]. 北京: 北京工业大学, 2019.
	CUI J X. Research and verification of rapid source traceability and emergency assessment technology for sudden air pollution accidents[D]. Beijing: Beijing Industry University, 2019. (in Chinese)
[26]	CHANG J C, HANNA S R. Air quality model performance evaluation[J]. Meteorology & Atmospheric Physics, 2004, 87(1/2/3):167-196.
[27]	邹敬, 曲静原, 曹建主. 大气扩散模型验证与比对的工具和方法[J]. 辐射防护通讯, 2004, 24(5):15-20.
	ZOU J, QU J Y, CAO J Z. Tools and methods for verification and comparison of atmospheric diffusion models[J]. Radiation Protection Communication, 2004, 24(5):15-20. (in Chinese)
[28]	宁莎莎, 李璟涛, 张怀宇. 核事故应急大气扩散模型ARTM验证与评价[J]. 南方能源建设, 2020, 7(4):93-98.
	NING S S, LI J T, ZHANG H Y. Evaluation of nuclear accident atmospheric tadionuclide transport model ARTM[J]. Southern Energy Construction, 2020, 7(4): 93-98. (in Chinese)

方向	扩散方程	大气稳定度	γ	μ	下风距离/m
y	σ_y=γx^μ	D级	0.110726	0.929418	1~1000
			0.146669	0.888723	>1000
z	σ_z=γx^μ	D级	0.104634	0.826212	1~1000
			0.400167	0.632023	>1000

方向	扩散方程	大气稳定度	γ	μ	下风距离/m
y	σ_y=γx^μ	D级	0.110726	0.929418	1~1000
			0.146669	0.888723	>1000
z	σ_z=γx^μ	D级	0.104634	0.826212	1~1000
			0.400167	0.632023	>1000

y/m	x/m
y/m	250	500	750	1000	1250
100	31.95802862	387.08525	367.6454777	278.3340544	90.40847051
60	630.7893626	1121.3852	594.006858	366.1385291	261.0453579
20	575.6576425	773.066851	416.7973147	333.6252268	245.2681073
-20	0.416086777	36.60284873	227.8322835	209.8455332	181.8178239
-60	9.57555×10^-7	5.634278715	53.85226113	92.8115419	105.2333203
-100	6.8812×10^-15	0.057732329	6.802763206	13.05211306	47.75791929

y/m	x/m
y/m	250	500	750	1000	1250
100	31.95802862	387.08525	367.6454777	278.3340544	90.40847051
60	630.7893626	1121.3852	594.006858	366.1385291	261.0453579
20	575.6576425	773.066851	416.7973147	333.6252268	245.2681073
-20	0.416086777	36.60284873	227.8322835	209.8455332	181.8178239
-60	9.57555×10^-7	5.634278715	53.85226113	92.8115419	105.2333203
-100	6.8812×10^-15	0.057732329	6.802763206	13.05211306	47.75791929

运行次数	x/m	y/m	He/m	Q/(g·s^-1)	耗时/s	x相对误差/10^-4	Q相对误差/10^-5
1	30.01	50.00	10.00	4999783.77	236.204	3.33	4.32
2	29.99	50.00	10.00	5000205.93	238.033	3.33	4.12
3	30.01	50.00	10.00	4999787.93	270.030	3.33	4.24
4	30.01	50.00	10.00	4999782.37	201.632	3.33	4.35
5	30.01	50.00	10.00	4999769.90	264.663	3.33	4.60
6	30.01	50.00	10.00	4999820.43	177.936	3.33	3.59
7	30.01	50.00	10.00	4999805.10	279.241	3.33	3.90
8	30.01	50.00	10.00	4999780.06	264.850	3.33	4.40
9	29.99	50.00	10.00	5000217.31	190.394	3.33	4.35
10	30.01	50.00	10.00	4999758.12	323.281	3.33	4.84