Source strength inversion of PSO-IA under modified Gaussian models

doi:10.16265/j.cnki.issn1003-3033.2024.07.0146

Abstract

Abstract:

In order to improve the science and effectiveness of traceability and localization of hazardous gas leaks, determining the location and intensity of dangerous gas leaks is the key to emergency response to accidents. The Gaussian plume model was modified by analyzing the mass conservation law and improving the diffusion amplitude of the gas plume with an approximate Gaussian distribution. Additionally, a heuristic algorithm based on the principle of immunization—IA coupled with PSO—was proposed, and the PSO-IA algorithm was applied to source strength inversion. It is concluded that the modified Gaussian plume model has been verified by three classical algorithms (PS, GA and PSO), resulting in a prediction value error decreased by about 2%. PSO algorithm, which showed a better inversion effect, was selected for comparison with the PSO-IA algorithm. The PSO-IA algorithm has improved the effect of inverting source strength, with a localization error is 1.3 m, a source strength solving error of 0.8%, and a single computation time of less than 1 second. This enables fast and accurate positioning and estimation of source strength.

Key words: particle swarm optimization-immune algorithm(PSO-IA), modified Gaussian smoke plume model, source-strength inversion, hazardous gas leakage, solving accuracy

CLC Number:

X937

WAN Bangyin, KUAI Niansheng, HE Xiongyuan, PENG Minjun, DENG Limin. Source strength inversion of PSO-IA under modified Gaussian models[J]. China Safety Science Journal, 2024, 34(7): 132-138.

Figures/Tables 5

Fig.1

Fig.2

Table 1

Fig.3

Table 2

References 21

[1]	KEATS A, YEE E, LIEN F S. Bayesian inference for source determination with applications to a complex urban environment[J]. Atmospheric Environment, 2007, 41(3):465-479.
[2]	CHOW F K, KOSOVIC B, CHAN S. Source inversion for contaminant plume dispersion in urban environments using building-resolving simulations[J]. Journal of Applied Meteorology and Climatology, 2008, 48(6):1553-1572.
[3]	HAUPT S E, BEYER-LOUT A, LONG K J, et al. Assimilating concentration observations for transport and dispersion modeling in a meandering wind field[J]. Atmospheric Environment, 2009, 43(6):1329-1338.
[4]	HENG Xiaoping, CHEN Zengqiang. Back calculation of the strength and location of hazardous materials releases using the pattern search method[J]. Journal of Hazardous Materials, 2010, 183(1/2/3):474-481.
[5]	张建文, 刘茜, 魏利军. 危险化学品泄漏事故泄漏源强反演方法比较研究[J]. 中国安全科学学报, 2009, 19(2):165-171.
	ZHANG Jianwen, LIU Qian, WEI Lijun. Comparative study on the inverse-calculation methods for the intensity of leakage sources in chemical leakage accidents[J]. China Safety Science Journal, 2009, 19(2): 165-171.
[6]	张久凤, 姜春明, 王正, 等. 粒子群优化算法在源强反演问题中的应用研究[J]. 中国安全科学学报, 2010, 20(10):123-128.
	ZHANG Jiufeng, JIANG Chunming, WANG Zheng, et al. PSO algorithm for inverse-calculation of source intensity[J]. China Safety Science Journal, 2010 20(10):123-128.
[7]	SVEN-ERIK G, ERIK L. Atmospheric dispersion from elevated sources in an urban area comparison between tracer experiments and model calculations[J]. Journal of Climate and Applied Meteorology, 1984, 23(4):651-660.
[8]	刘畅, 苏腾, 周汝, 等. 修正高斯模型下气体泄漏源项信息反演研究[J]. 中国安全科学学报, 2022, 32(7):98-104. doi: 10.16265/j.cnki.issn1003-3033.2022.07.0741
	LIU Chang, SU Teng, ZHOU Ru, et al. Investigation on inverse-caleulation of leakage source information of gas based on modified Gaussian model[J]. China Safety Science Journal ̧ 2022, 32(7):98-104.
[9]	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 Zengqiang. Research on source identification approaches for hazardous chemical releases[D]. Beijing: Beijing University of Chemical Technology, 2013.
[10]	HJ 169—2018,建设项目环境风险评价技术导则[S].
	HJ 169-2018,Technical guidelines for environmental risk assessment on projects[S].
[11]	GIFFORD J F A. Use of routine meteorological observations for estimating atmospheric dispersion[J]. Nuclear Safety, 1961, 2: 47-51.
[12]	胡爽, 汤亚玲. 基于改进高斯烟羽模型的二氧化氮泄漏模拟分析[J]. 重庆工商大学学报:自然科学版, 2023, 40(1):22-27.
	HU Shuang, TANG Yaling. Simulation analysis of nitrogen dioxide leakage based on improved Gaussian plume model[J]. Journal of Chongqing Technology and Business University:Natural Science Edition, 2023, 40(1): 22-27.
[13]	HE Peng, ZHENG Bohong, ZHENG Jian. Urban PM2.5 diffusion analysis based on the improved gaussian smoke plume model and support vector machine[J]. Aerosol and Air Quality Research, 2018, 18(12):3177-3186.
[14]	WILSON J D, FERRANDINO F J, THURTELL G W. A relationship between deposition velocity and trajectory reflection probability for use in stochastic lagrangian dispersion models[J]. Agricultural and Forest Meteorology, 1989, 47(2/3/4):139-154.
[15]	李云云. 高斯烟羽模型的改进及在危化品泄漏事故模拟中的应用[D]. 广州: 广州大学, 2013.
[16]	付金宇, 李颖. 基于高斯烟羽模型的船舶尾气扩散研究[J]. 海洋通报. 2018, 37(2):235-240.
	FU Jinyu, LI Ying. Study on ship's exhaust-gas diffusion based on Gaussian plume model[J]. Marine Science Bulletin, 2018, 37(2):235-240.
[17]	THOMSON L C, HIRST B, GIBSON G, et al. An improved algorithm for locating a gas source using inverse methods[J]. Atmospheric Environment, 2007, 41(6): 1128-1134.
[18]	孙宁. 人工免疫优化算法及其应用研究[D]. 哈尔滨: 哈尔滨工业大学, 2006.
	SUN Ning. Artificial immune optimization algorithm and applications[D]. Harbin:Harbin Institute of Technology, 2006.
[19]	吕红芳, 王涛, 嵇月强, 等. 基于免疫粒子群算法的PID参数优化研究[J]. 中国工程机械学报, 2022, 20(3):194-198.
	LYU Hongfang, WANG Tao, JI Yueqiang, et al. Optimization of PlD control parameters based on immune particle swarm optimization[J]. Chinese Journal of Construction Machinery, 2022, 20(3):194-198.
[20]	GB/T37243—2019,危险化学品生产装置和储存设施外部安全防护距离确定方法[S].
	GB/T37243-2019,Determination method of external safety distance for hazardous chemicals production units and storage installations[S].
[21]	沈泽亚, 郎建垒, 程水源, 等. 典型耦合优化算法在源项反演中的对比研究[J]. 中国环境科学, 2019, 39(8):3207-3214.
	SHEN Zeya, LANG Jianlei, CHENG Shuiyuan, et al. Comparative and study on the application of typical hybridalgorithms in source parameter inversions[J]. China Environmental Science, 2019, 39(8):3207-3214.

对比模型	经典算法	α	x值/m	y值/m	Q/(kg·s^-1)	适应度值 (方差和)	收敛迭代次数
高斯模型	PS	0.02	242.3	44.2	1.255	1.844×10^-4	401
	GA		276.4	42.2	1.249	1.832×10^-4	120
	PSO		252.2	42.6	1.223	1.826×10^-4	102
修正高斯模型	PS		248.5	45.1	1.216	1.806×10^-4	400
	GA		250.3	43.7	1.208	1.795×10^-4	126
	PSO		265.9	43	1.071	1.779×10^-4	113

算法	种群 N	迭代次数	真实坐标/m	反算均值坐标/m	偏离d/ m	θ真实值/ (kg·s^-1)	反算Q均值/ (kg·s^-1)	相对误差/%	计算总时间/s	平均单次时间/s
PSO	30	500	260,45.9	253.7,40.4	8.4	1.127	1.191	5.7	44.7	0.45
	50	500		265.6,42.8	6.4		1.076	4.5	70.9	0.71
	100	500		267.8,47.9	8.1		1.066	5.4	139.2	1.39
PSO-IA	30	500		255.2,41.6	6.4		1.099	2.5	45	0.45
	50	500		261.2,46.5	1.3		1.118	0.8	71.2	0.71
	100	500		262.5,43.1	3.8		1.137	0.9	140.8	1.41