文献类型：期刊文献

英文题名：Data-driven source term estimation of hazardous gas leakages under variable meteorological conditions

作者：Ni, Chuantao[1];Lang, Ziqiang[1,2];Wang, Bing[1];Li, Ang[1];Cao, Chenxi[1];Du, Wenli[1];Qian, Feng[1]

机构：[1]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[2]Univ Sheffield, Dept Automat Control & Syst Engn, Sheffield S1 3JD, England

年份：2025

卷号：94

外文期刊名：JOURNAL OF LOSS PREVENTION IN THE PROCESS INDUSTRIES

收录：;EI(收录号：20244917484384);WOS:【SCI-EXPANDED(收录号:WOS:001374697200001)】；

基金：The work was supported by the National Natural Science Foundation of China [Basic Science Center Program 61988101, Grant 62303186, and Grant 62203173] .

语种：英文

外文关键词：Source term estimation; Multi-sensor data-driven; Radial basis neural network; Variable meteorological conditions; Hazardous gas leakage

摘要：Source term estimation (STE) of hazardous gas leakages in chemical industrial parks (CIPs) is important for addressing environmental pollution and improving engineering safety and reliability. For this purpose, some least squares-based STE methods have recently been developed, which performs the real-time STE using an offline determined response matrix that represents the relationship between the sensor measurements and strengths of hazardous gas leakages. However, these methods require the number and locations of potential hazardous gas leakage sources are known a priori, which is difficult in many practical applications. To resolve this issue, in the present study, a novel multi-sensor data-driven STE (MSDD-STE) approach is proposed, which overcomes the difficulties with existing least squares-based STE methods and can, for the first time, address the MSDD-STE problems in complicated scenarios where meteorological conditions such as wind directions change over a considerable range. A detailed analysis is introduced to evaluate the performance of the proposed approach. The effectiveness of the proposed approach is verified by comprehensive numerical simulation studies.