~通电导线故障产生的迸溅熔珠易引燃木质可燃物造成火灾，尤其在城市-森林交界域，严重威胁着社会公共安全。与大量研究的传统火灾相比，此类火灾的引燃方式有很大不同。同时，对电气火灾调查而言，熔珠是证明电气故障发生的重要证据，如果可以确定熔珠具有引燃能力，并且证明熔珠为“起因熔珠”，即可为火灾原因的辨识提供可靠依据。但是由于迸溅熔珠的产生有多种诱因，且熔珠痕迹特征受火灾影响大，致使火灾调查原因认定困难，熔珠痕迹特征难以定量化。因此，本文进行了电弧迸溅熔珠引燃能力及痕迹特征辨识方法研究，实验模拟短路故障产生的迸溅熔珠引燃纤维素全过程（包括三个阶段：熔珠产生、飞行、着床引燃），建立迸溅熔珠引燃能力预测模型；观测不同引燃现象、引燃行为，并通过熔珠引燃纤维素的数值模拟，揭示不同引燃现象的发生机理；进而开展不同引燃现象对熔珠痕迹特征的影响研究；最后，建立基于熔珠痕迹特征的引燃结果辨识模型。主要工作和成果如下：
通过“电气火灾故障模拟及痕迹制备装置”，采用一种改进的熔珠制备方法，产生不同电弧能量的迸溅熔珠，并以木质可燃物的主要成分纤维素作为燃料。实验模拟电弧迸溅熔珠引燃纤维素全过程，通过统计分析，发现熔珠直径和数量之间并非彼此独立，存在电弧能量临界值域100–150 J，使得产生熔珠的直径和数量适中，容易导致引燃。同时发现，熔珠直径临界域1–3 mm对引燃起决定性作用。根据引燃临界条件，划分了熔珠引燃危险性临界值域：引燃区域、潜在引燃区域、不引燃区域。采用Logistic回归方法，建立了熔珠引燃能力预测模型，确定了引燃能力与电弧能量和直径的关系，发现熔珠直径的变化对引燃概率的影响更加敏感，尤其当电弧能量大于175 J条件下。模型从熔珠产生阶段出发，提出了基于标定物热惯性参数的电弧熔珠引燃能力预测方法。
基于高速摄像机记录的熔珠着床引燃阶段，观测到三种不同的实验现象：引燃、阴燃、不燃。通过定量参数化研究，发现熔珠着床前温度和直径直接影响熔珠所具有的体积能，但体积能并不能直接决定最终的引燃条件，反而，熔珠动能和燃料床密度对引燃条件的影响更大，其主要影响熔珠的着床引燃方式、着床深度。通过数值模拟研究，考虑了熔珠点燃纤维素的两步反应，同时，考虑了材料表面上方的自然对流、扩散和热辐射，该模型确定了熔珠所需的最低点火温度1300 K，熔珠恰好全部嵌入燃料床，引燃时间最短，数值模型与实验结果吻合较好；模型得到了不同引燃现象高温点及气体浓度运移规律，发现不同引燃现象发生位置处在放热量、挥发物浓度和氧气浓度充足的地方，从而揭示了不同引燃现象发生机理：不燃、纤维素内部炭化阴燃、阴燃点燃挥发物、明火直接点燃挥发物。
根据GB16840.4熔痕的判别标准，将不同引燃现象熔珠微观特征进行定性分类，得到引燃、阴燃熔痕可分为3类，不燃熔痕可分为4类；采用Image-Pro Plus图像处理软件，定量研究不同引燃现象熔珠的金相组织、气孔分布规律，得出了不同引燃现象熔珠金相组织存在差异，气孔随着床深度的增加而明显增多。确定了6个指标参数：熔珠晶粒面积、平均直径、圆度，和气孔面积、平均直径、圆度，可作为判别因子，建立了熔珠微观特征参数Fisher-Bayes判别模型，得出了熔珠痕迹特征与不同引燃现象之间的关系。
针对熔珠引燃结果和熔珠痕迹特征之间的复杂非线性关系，采用了支持向量机（SVM）学习算法，发现SVM算法分类结果的精度受其超参数影响很大，其超参数之间存在一个最佳匹配与组合关系。因此，提出了采用混合模拟退火粒子群（SAPSO）算法对SVM超参数进行寻优，经过SAPSO优化模型分类的准确率、敏感性和特异性均高于98%。主成分分析（PCA）降维方法的精度与其主成分提取信息累积仅包含整个原始数据信息的有关。但对于模型的泛化性，则PCA-SAPSO-SVM模型的普适性较好，模型精度可达91%，是一种简单准确、稳定可靠的基于熔珠痕迹特征的引燃结果辨识模型。最终，建立较完善的迸溅熔珠引燃木质可燃物火灾原因辨识方法。
研究成果为电气火灾物证提取与鉴定提供理论基础和方法支撑，并为避免城市-森林交界域火灾提供安全指南，对于制定有效的消防安全管理和电气火灾调查标准具有实际意义。
 

~The splashing beads generated by the electrical failure can easily ignite the wood combustibles and cause fires, especially in wildland-urban interface areas, which seriously threatens social and public safety. Compared with traditional fires commonly, the ignition methods of such fires are highly different. Meanwhile, for investigating electrical fires, the pattern characteristics of splashing beads are the key evidence to prove the occurrence of electrical faults. If it can be determined that the beads have the ignition ability, and it is proved that the beads are “cause beads”, it can provide strong evidence for the determination of the fire cause. However, since there are various incentives for the generation of splashed beads, and the pattern characteristics of the beads are greatly affected by the fire. Thus, it is difficult to identify the cause of fire investigation and the pattern characteristics are difficult to quantify. Therefore, this paper has carried out research on ignition ability and pattern characteristics identification method of arc splashing beads. The experiment simulates the whole process of igniting wood combustibles by arc splashing beads generated by short-circuit faults (including 3 stages: beads generation, flight, and ignition), and a prediction model of the ignition ability of splash beads is established; Observed the characteristics of different ignition behaviors and ignition phenomena, and, through numerical simulation, the mechanism of different ignition phenomena of splashed beads is revealed; Furthermore, the influence of different ignition phenomena on the characteristics of bead pattern was studied; Finally, an identification model of ignition results based on the characteristics of bead patterns is established. The main work and results are as follows:
Through the “Electrical fire fault simulation and beads pattern device”, an improved molten bead preparation method was used to generate splashed molten beads of different arc energies, and cellulose, a well-characterized wood combustible material, was used as a fuel. The experiment simulates the whole process of arc splashing beads to ignite cellulose fuel. Through the statistical analysis, it is found that the diameter and number of molten beads are not independent of each other, and there is a critical value range of arc energy of 100–150 J, which makes the diameter and number of beads moderate, which is conducive to the occurrence of ignition. At the same time, the critical region of the bead diameter of 1–3 mm plays a decisive role in ignition. According to the ignition critical conditions, the critical range of ignition danger is divided: ignition region, potential ignition region, and no-ignition region. Logistic regression method is used to establish a predictive model for the ignition capacity of the bead, and the relationship between the ignition capacity and the arc energy and diameter is determined. The change of the diameter of the bead is more sensitive to the influence of the ignition probability, especially when the arc energy is greater than 175 J. The model is based on the bead generation stage, and a method for predicting the ignition ability of the arc bead based on the thermal inertia parameter of the calibration object is proposed.
Based on the experimental phenomena recorded by the high-speed camera, the different ignition situations caused by the beads are divided into three ignition phenomena: Flaming ignition, smoldering ignition, and no-ignition. Through quantitative parameterization research, it is found that the temperature before implantation and the diameter of the bead directly affect the volume energy of the bead, but the volume energy does not directly determine the ignition phenomenon. More importantly, the ignition method and the implantation process have a greater impact on the ignition phenomenon. Based on the experimental observations, a mathematical model was developed, and the two-step reaction of the ignition of cellulosic by arc beads was considered. Meanwhile, the natural convection, diffusion and thermal radiation above the surface of the material were premeditated. The model determines that the minimum ignition temperature required by the beads is 1300 K. The beads are just all embedded in the fuel bed, and the ignition time is the shortest. The numerical model is in good agreement with the experimental results. Through numerical simulation research, the high temperature point and gas concentration migration law of no-ignition, internal carbonization smoldering of cellulose, smoldering volatiles, and direct ignition of flames are obtained. Ignition phenomenon was limited by heat conduction, accumulation of heat release, volatile substance, and oxygen concentration. The model can reasonably reveal the occurrence mechanism of different ignition phenomena, including no-ignition, carbonized smoldering inside cellulose, smoldering ignites volatiles, and flaming ignition volatiles directly.
According to the standard of GB16840.4, the different ignition phenomena beads are qualitatively classified, ignition and smoldering pattern characteristics can be divided into 3 categories, and non-ignition pattern characteristics can be divided into 4 categories. Using Image-Pro Plus image processing software, quantitatively study the metallographic structure and pore distribution of different ignition types of the beads. It is concluded that there are differences in the microscopic morphology of the beads with different ignition phenomena, and the porosity increases significantly with the depth of the fuel bed. In addition, obtained the crystal grain area, average grain diameter, grain roundness, and pore area of the melt beads, the average diameter, and the roundness of stomata can be used as discriminating factors. The Fisher-Bayes discriminant model of the microscopic characteristic parameters of the beads is established, and the relationship between the typical discriminant factors and different ignition phenomena is determined.
Aiming at the complex nonlinear relationship between the ignition result of the beads pattern characteristics, a support vector machine (SVM) learning algorithm is introduced. It is found that the accuracy of the classification results of the SVM method is greatly affected by its hyperparameters, and there is an optimal matching and combination relationship between the hyperparameters. The hybrid simulated annealing particle swarm optimization (SAPSO) algorithm is proposed to optimize the SVM hyperparameters. The accuracy, sensitivity, and specificity of the SAPSO optimization model classification are all up to 98%. The accuracy of the principal component analysis (PCA) dimensionality reduction method is related to the fact that the accumulation of principal component extraction information only contains the entire original data information. But for the popularization and application of the model, the generalization of the PCA-SAPSO-SVM model is better, the accuracy can reach 91% of the PCA-SAPSO-SVM model has been greatly improved, which is a simple, accurate, stable, and reliable model to identify the ignition result based on the pattern characteristics of the bead. Finally, a relatively complete fire cause identification method for igniting wood combustibles by splashing beads is established.
The results provide a theoretical basis and method support for the extraction and identification of electrical fire evidence, and provide safe guidelines for avoiding wildland-urban interface fires, which are of practical significance for the formulation of effective fire safety management and electrical fire investigation standards.
 

[1] 应急管理部消防救援局. 数据统计. https://www.119.gov.cn/gongkai/sjtj

[2] 应急管理部消防救援局. 2021年消防接处警创新高，扑救火灾74.5万起. https://www.119.gov.cn/article/46TiYamnnrs.

[3] Rein G. SFPE handbook of fire protection engineering (Fifth Edition): Smoldering Combustion [M]. Springer New Yorker Heidelberg Dordrecht London, 2016.

[4] Ramljak I, Majstrovic M, Sutloviec E. Statistical analysis of particles conductor clashing [C]. IEEE, 2014: 638–643.

[5] Park S H, Lim S J, Cha M S, et al. Effect of AC electric field on flame spread in electrical wire: Variation in polyethylene insulation thickness and di-electrophoresis phenomenon [J]. Combustion and Flame, 2019, 202: 107–118.

[6] Koo E, Pagni P J, Weise D R, et al. Firebrands and spotting ignition in large-scale fires [J]. International Journal of Wildland Fire, 2010, 19 (7): 818–843.

[7] Fernandez-Pello A C, Lautenberger C, Rich D, et al. Spot fire ignition of natural fuel beds by hot metal particles, embers, and sparks [J]. Combustion science and technology, 2015, 187 (1–2): 269–295.

[8] NFPA, NFPA Standard 51B: Standard for Fire Prevention During Welding, Cutting, and Other Hot Work[J]. National Fire Protection Association, Quincy, MA, 2017.

[9] Mell W E, Manzello S L, Maranghides A, et al. The wildland–urban interface fire problem–current approaches and research needs [J]. International Journal of Wildland Fire, 2010, 19 (2): 238–251.

[10] 新华社. 四川凉山西昌“3·30” 森林火灾事件调查结果公布. 消防界(电子版) , 2021, 02. DOI:10.16859/j.cnki.cn12-9204/tu.2021.02.016.

[11] Maranghides A, Mell W. A case study of a community affected by the Witch and Guejito wildland fires [J]. Fire technology, 2011, 47 (2): 379–420.

[12] Babrauskas V. SFPE handbook of fire protection engineering (Fifth Edition): Electrical Fires [M]. Springer New Yorker Heidelberg Dordrecht London, 2016.

[13] Yuan C, Liu K, Amyotte P, et al. Electric spark ignition sensitivity of nano and micro Ti powder layers in the presence of inert nano TiO2 powder [J]. Journal of Loss Prevention in the Process Industries, 2017, 46: 84–93.

[14] Babrauskas V. How do electrical wiring faults lead to structure ignitions [C]. Process Fire and Materials 2001 Conference, 2001: 39–51.

[15] Uber C, Shekhar R, Felgner A, et al. Experimental investigation of low-voltage spark ignition caused by separating electrodes [J]. Journal of Loss Prevention in the Process Industries, 2017, 49: 822–831.

[16] von Pidoll U. The ignition of clouds of sprays, powders and fibers by flames and electric sparks [J]. Journal of Loss Prevention in the Process Industries, 2001, 14 (2): 103–109.

[17] Liu K H, Shih Y H, Chen G J, et al. Microstructural study on molten marks of fire-causing copper wires[J]. Materials, 2015, 8 (6): 3776–3790.

[18] Bu Y, Yuan C, Amyotte P, et al. Ignition hazard of non-metallic dust clouds exposed to hotspots versus electrical sparks [J]. Journal of hazardous materials, 2019, 365: 895–904.

[19] Iwashita T, Hagimoto Y, Sugawa O. Characterization of arc beads on energized conductors exposed to radiant heat [J]. Fire and Materials, 2017, 41 (8): 1072–1078.

[20] Novak C J, Stoliarov S I, Keller M R, et al. An analysis of heat flux induced arc formation in a residential electrical cable [J]. Fire safety journal, 2013, 55: 61–68.

[21] Babrauskas V. Research on electrical fires: the state of the art [J]. Fire Safety Science, 2008, 9:3–18.

[22] Gorbett G E, Meacham B J, Wood C B, et al. Use of damage in fire investigation: a review of fire patterns analysis, research and future direction [J]. Fire Science Reviews, 2015, 4 (1): 1–35.

[23] Tinsley A, Gorbett G. Fire investigation origin determination survey [J]. Fire and Arson Investigator Journal of the International Association of Arson Investigators, 2013, 63: 24–40.

[24] Babrauskas V. NFPA 921: Guide for Fire and Explosion Investigations (2017 Edition): Electrical Fires [M]. National Fire Protection Association, Quincy, MA, 2017.

[25] Urban J L. Spot ignition of natural fuels by hot metal particles [D]. University of California, Berkeley, 2017.

[26] 中华人民共和国中央人民政府. 国务院安委会启动为期3年的电气火灾综合治理行动. http://www.gov.cn/xinwen/2017-05/03/content_5190668.htm

[27] Hutchison V. Residential electrical fire problem: the data landscape [M]. Fire Protection Research Foundation, 2018.

[28] Babrauskas V, Simonson M. Fire behaviour of plastic parts in electrical appliances-standards versus required fire safety objectives [J]. Fire and Materials: An International Journal, 2007, 31 (1): 83–96.

[29] 何洪源. 火灾现场调查与火灾物证分析[M]. 北京：中国人民公安大学出版社，2010.

[30] Babrauskas V. Electrical fires: research needed to improve safety [J]. Journal of Fire Protection Engineering, 2010, 46: 20–30.

[31] Hagimoto Y, Watanabe N, Okamoto K. A Short Circuit as an Ignition Source of Fire [C]. In Inter flam 2007, Inter science Communications Ltd, London, 2007, 1555–1560.

[32] Babrauskas V. Mechanisms and modes for ignition of low-voltage, PVC-insulated electrotechnical products [J]. Fire and Materials, 2006, 30: 150–174.

[33] Coldham D, Czerwinski A, Marxsen T. Probability of bushfire ignition from electric arc faults [J]. HRL Technology Pty Ltd., Melbourne, VIC., Australia, Tech. Rep. HRL, 2010:195, 2011.

[34] Novak C J, Stoliarov S I, Keller M R, et al. An analysis of heat flux induced arc formation in a residential electrical cable [J]. Fire Safety Journal, 2013, 55: 61–68.

[35] Iwashita T, Hagimoto Y, Sugawa O. Characterization of arc beads on energized conductors exposed to radiant heat [J]. Fire and Materials, 2017, 41 (8): 1072–1078.

[36] Iwashita T, Keller M R, Hagimoto Y, et al. Leakage currents precede short circuits in PVC-insulated cable when exposed to external radiant heat [J]. Fire and Materials, 2017, 41(4): 339–348.

[37] Kanokbannakorn W, Hongesombut K, Teerakawanich N, et al. Arc flash hazard in distribution system with distributed generation [J]. Procedia Computer Science, 2016, 86: 377–380.

[38] Kumpulainen L, Hussain G A, Rival M, et al. Aspects of arc-flash protection and prediction [J]. Electric Power Systems Research, 2014, 116: 77–86.

[39] Hagimoto, Y, Watanabe, N., Okamoto. Arcing faults on PVC covered electrical cords [C]. Proc. 1st Con! Of the Assn. of Korean Japanese Sa fely Engineering Society, Kyongju, Korea, 1999, 221–224.

[40] Franklin F F. Circuit breakers: The myth of safety [J]. Professional Safety, 1990, 35(6): 28.

[41] British Standards Institution. IEC 62271-200, High-voltage switchgear and controlgear-Part 200: AC metal-enclosed switchgear and controlgear for rated voltages above 1 kV and up toand including 52 kV [M]. Edition 2.0, 2011.

[42] Arvola J, Dahl S, Virtala T. Improving medium voltage switchgear protection in compensated distribution networks [C]. CIRED 2013 Conference, 2013: 10–13.

[43] Cressault Y, Murphy A B, Teulet P, et al. Thermal plasma properties for Ar-Cu, Ar-Fe and Ar-Al mixtures used in welding plasmas processes: II. Transport coefficients at atmospheric pressure [J]. Journal of Physics D: Applied Physics, 2013, 46: 415207.

[44] Babrauskas V. Electric arc explosions-A review [J]. Fire Safety Journal, 2017, 89: 7–15.

[45] Khakpour A, Franke S, Uhrlandt D, et al. Electrical arc model based on physical parameters and power calculation [J]. IEEE Transactions on Plasma Science, 2015, 43 (8): 2721–2729.

[46] Sawicki A. Problems of modeling an electrical arc with variable geometric dimensions [J]. Przeglad Elektrotechniczny, 2013, 89 (2): 270–275.

[47] Khakpour A, Franke S, Gortschakow S, et al. An improved arc model based on the arc diameter [J]. IEEE Transactions on Power Delivery, 2015, 31: 1335–1341.

[48] Walter M M, Franck C M. Improved method for direct black-box arc parameter determination and model validation [J]. IEEE Transactions on Power Delivery, 2014, 29 (2): 580–588.

[49] Khakpour A, Uhrlandt D, Methling R P, et al. Impact of temperature changing on voltage and power of an electric arc [J]. Electric Power Systems Research, 2017, 143: 73–83.

[50] Varol H, Cashell K A. Numerical modelling of high strength steel beams at elevated temperature [J]. Fire Safety Journal, 2017, 89: 41–50.

[51] Chen K C, Warne L K, Jorgenson R E, et al. TATB sensitivity to shocks from electrical arcs [J]. Propellants, Explosives, Pyrotechnics, 2019, 44 (8): 1000–1009.

[52] Senk G, Jakubove I, Laznickova I. Analysis of intensively blasted electric arc burning in the arc heater’s anode channel [J]. Acta Polytechnica, 2016, 56:395–401.

[53] Wan G, Dong Q, Zhi J, et al. Analysis on electrical and thermal conduction of carbon fiber composites under lightning based on electrical-thermal-chemical coupling and arc heating models [J]. Composite Structures, 2019, 229: 111486.

[54] Wu C, Zhao Q, Li Na, et al. Influence of fabrication technology on arc erosion of Ag/10SnO2 electrical contact materials [J]. Journal of Alloys and Compounds, 2018, 766: 161–177.

[55] Chen S, Zhang R, Jiang F, et al. Experimental study on electrical property of arc column in plasma arc welding [J]. Journal of Manufacturing Processes, 2018, 31: 823–832.

[56] Zhou YX, Xue YL, Zhou K. Failure analysis of arc ablated tungsten-copper electrical contacts [J]. Vacuum, 2019, 164: 390–395.

[57] Doddapaneni V, Saleemi M, Ye F, et al. Engineered PMMA-ZnO nanocomposites for improving the electric arc interruption capability in electrical switching applications: Unprecedented experimental insights [J]. Composites Science and Technology, 2017, 141: 113–119.

[58] Calderon-Mendoza E, Schweitzer P, Weber S. Kalman filter and a fuzzy logic processor for series arcing fault detection in a home electrical network [J]. International Journal of Electrical Power and Energy Systems, 2019, 107: 251–263.

[59] Lin Z, Fan S, Liu M, et al. Excellent anti-arc erosion performance and corresponding mechanisms of a nickel-belt-reinforced silver-based electrical contact material [J]. Journal of Alloys and Compounds, 2019, 788: 163–171.

[60] Wen X, Yuwen F, Ding Z, et al. Electric arc-induced damage on electroless Ag film using ionic liquid as a lubricant under sliding electrical contact [J]. Tribology International, 2019, 135: 269–276.

[61] Zhu S, Liu Y, Tian B, et al. Arc erosion behavior and mechanism of Cu/Cr20 electrical contact material [J]. Vacuum, 2017, 143: 129–137.

[62] Chabrerie J P, Devautour J, Teste P. A numerical model for thermal processes in an electrode submitted to an arc in air and its experimental verification [J]. IEEE transactions on components, hybrids, and manufacturing technology, 1993, 16 (4): 449–455.

[63] Sun M, Wang Qi, Lindmayer M. The model of interaction between arc and AgMeO contact material [J]. IEEE Transactions on Components, Packaging, Manufacture Technology, 1994, 17: 490–494.

[64] Wu X, Li Z B, Tian Y, et al. Investigate on the simulation of black-box arc model [C].2011 1st International Conference on Electric Power Equipment-Switching Technology. IEEE, 2011: 629–636.

[65] Chen Z, Koichiro S. Effect of arc behavior on material transfer: A Review [J]. IEEE Transaction on Components, Packaging and Manufacturing Technology-part A. 1998, 21: 310–322.

[66] Chung H H, Lee R T, Chiou Y C. Erosion mechanism of silver in a single arc discharge across a static gap [J].  IEE Proceedings Science Measurement and Technology, 2002, 149: 172–180.

[67] Weaver P M, Pechrach K, McBride J W. The energetics of gas flow and contact erosion during short circuit arcing. IEEE Transaction on Components and Packaging Technology. 2004, 27: 51–56.

[68] 郑梦笛, 吴细秀, 闫格. 基于新型故障电弧模型的电弧能量特性分析[J]. 中国电力, 2015, 48 (11): 49–53.

[69] 吴细秀. 开关电器触头材料喷溅侵蚀模型研究及其试验[D]. 华中科技大学, 2005.

[70] Wang Y. Study on the splashing short-circuited beads’ characteristics in fire scenes by density methods [J]. Procedia Engineering, 2016, 135: 555–562.

[71] Ogale S B, Shinde S R, Karve P A, et al. Impact-induced splash and spill in a quasi-confined granular medium [J]. Physica A: Statistical Mechanics and its Applications, 2006, 363 (2): 187–197.

[72] Pleasance G E, Hart J A. An examination of particles from conductors clashingas possible source of bushfire ignition. State Electricity Commission of Victoria (SEC), Victoria, Australia, Research and Development Department, Report FM-1, 1977.

[73] Pleasance G E, Hart J A. An examination of particles from conductor clashes as possible sources of bushfire ignition [J]. Laboratory report, SECV Victoria, 1977.

[74] Joynt R. The possibility of fires being caused by copper conductor clashing. Tech. rep. Victoria, Australia: State Electricity Commission of Victoria, 1983.

[75] Coldham D. Bushfire ignition from electric faults: A review of technical literature. Tech. rep. Victoria, Australia: Energy Safe Victoria, 2011.

[76] Elis S, Matislav M. Statistical analysis of particles of conductor clashing [J]. Energy Conservation, 2014, 11 (3): 638–643.

[77] Mikkelsen K. An experimental investigation of ignition propensity of hot work processes in the nuclear industry [J]. Zeitschrift Fur Naturforschung A, 2014.

[78] Fernandez-Pello A C. Wildland fire spot ignition by sparks and firebrands [J]. Fire Safety Journal, 2017, 91: 2–10.

[79] Baum HR, McCaffrey BJ. Fire induced flow field-theory and experiments fire safety science [C]. Proceedings of the Second International Symposium, Washington DC, 1989, 129–148.

[80] McGrattan K B, Baum H R, Rehn R G. Smoke plumes from large fires, UJNR panel on Fire Research, NIST, Gaithersburgh, MD, 1995.

[81] Quintiere J G, Grove B S. A unified analysis for fire plumes [C]. Symposium (International) on Combustion. Elsevier, 1998, 27 (2): 2757–2766.

[82] Woycheese J P, Pagni P J, Liepmann D. Brand propagation from large-scale fires [J]. Journal of Fire Protection Engineering, 1999, 10 (2): 32–44.

[83] Huang H, Ooka R, Kato S, et al. A numerical study of firebrands scattering in urban fire based on CFD and firebrands aerodynamics measurements [J]. Journal of fire sciences, 2007, 25 (4): 355–378.

[84] Himoto K, Mauyama T, Tanaka T. A study on the brand spotting in urban fires-LES analysis on the scattering of square disks in a turbulent boundary layer [C]. Proceedings of the 10th Inter flammation, 2004, 1039–1050.

[85] Satoh K, Kuwahara K, Yang K T. A numerical study of forest fire progression and fire suppression by aerial fire fighting [C]. ASME International Mechanical Engineering Congress and Exposition. 2004, 4708: 79-86.

[86] Tarifa C S, Del Notario P P, Moreno F G. On the flight paths and lifetimes of burning particles of wood, Symposium (International) on Combustion, 1965, 10: 1021–1037.

[87] Tarifa CS. Open fires and transport of firebrands. US department of agriculture forest service, 4th annual report of grant FG-SP-114. Madrid, 1965.

[88] Tarifa C S, Notario del P P, Moreno F G, et al. Transport and combustion of firebrands. Institute Nacional de Tecnica Aeroespacial, Madrid, Spain, 1967.

[89] Lee S L, Hellman J M. Study of firebrand trajectories in a turbulent swirling natural convection plume [J]. Combustion and Flame, 1969, 13 (6): 645–655.

[90] Lee S L, Hellman J M. Firebrand trajectory study using an empirical velocity-dependent burning law [J]. Combustion and Flame, 1970, 15 (3): 265–274.

[91] Mills A. Trajectories of sparks from arcing aluminum power cables [J]. Fire Technology, 1984, 20: 5–14.

[92] Tse S D, Fernandez-Pello A C. On the flight paths of metal particles and embers generated by power lines in high winds-a potential source of wildland fires [J]. Fire Safety Journal, 1998, 30 (4): 333–356.

[93] Psarros E G, Polykrati A D, Karagiannopoulos C G, et al. A model for calculating the temperature of aluminium particles ejected from overhead low-voltage lines owing to a short-circuit [J]. International Journal of Wildland Fire, 2009, 18 (6):722–726.

[94] Liu Y, Urban J L, Xu C, et al. Temperature and motion tracking of metal spark sprays [J]. Fire Technology, 2019, 55 (6): 2143–2169.

[95] Sun J, Hu L, Zhang Y. A review on research of fire dynamics in high-rise buildings [J]. Theoretical Applied Mechanics Letters, 2013, 3 (4): 042001.

[96] Shi L, Chew M Y L. A review of fire processes modeling of combustible materials under external heat flux [J]. Fuel, 2013, 106: 30–50.

[97] 卢志刚, 陈伟红. 电焊熔珠对棉布和聚氨酯泡沫的引燃能力研究[J]. 火灾科学, 2009, 18(01): 15–19.

[98] 赵艳红, 赵海龙, 赵志刚. 电焊渣对外墙保温材料引燃能力的试验研究[J]. 武警学院学报, 2016, 32(08): 25–31.

[99] 李青. 聚氨酯保温板引燃特性试验研究[J]. 武警学院报, 2015, 31(4): 24–27.

[100] 杨玖玲, 陈海翔. 高温金属颗粒作用下可燃物的点燃机理[J]. 燃烧科学与技术, 2014, 20(4): 329–334.

[101] Yang J, Wang S, Chen H. Effect of interface thermal resistance on ignition of reactive material by a hot particle [J]. International Journal of Heat and Mass Transfer, 2016, 97:146–156.

[102] Yang J , Rein G , Chen H , et al. Smoldering propensity in upholstered furniture: Effects of mockup configuration and foam thickness [J]. Applied Thermal Engineering, 2020, 181:115873.

[103] Yang J , Liu N , Chen H , et al. Smoldering and spontaneous transition to flaming over horizontal cellulosic insulation [J]. Proceedings of the Combustion Institute, 2019, 37 (3):4073–4081.

[104] Yang J, Liu N, Chen H, et al. Effects of atmospheric oxygen on horizontal peat smoldering fires: Experimental and numerical study [J]. Proceedings of the Combustion Institute, 2019, 37 (3): 4063-4071.

[105] 王苏盼. 飞火颗粒点燃的实验及机理研究[D]. 中国科学技术大学, 2016.

[106] Wang S, Chen H, Liu N. Ignition of expandable polystyrene foam by a hot particle: An experimental and numerical study [J]. Journal of Hazardous Materials, 2015, 283:536–543.

[107] Wang S, Chen H, Liu N, et al. Ignition of low-density expandable polystyrene foam by a hot particle [J]. Combustion and Flame, 2015, 162: 4112–4118.

[108] Song J, Wang S, Chen H. Safety distance for preventing hot particle ignition of building insulation materials [J]. Theoretical and Applied Mechanics Letters, 2014, 4 (3): 034005.

[109] Wang S, Chen H, Zhang L. Thermal decomposition kinetics of rigid polyurethane foam and ignition risk by a hot particle [J]. Journal of Applied Polymer Science, 2014, 131 (4).

[110] Wang Q, Liu K, Wang S. Effect of porosity on ignition and burning behavior of cellulose materials [J]. Fuel, 2022, 322: 124158.

[111] Wang S, Zhang Y, Huang X. Ignition of EPS foam by a hot moving hollow particle: Threshold, auto-ignition, and fire point [J]. Combustion and Flame, 2021, 232, 111524.

[112] Wang S, Huang X, Chen H, et al. Interaction between flaming and smouldering in hot-particle ignition of forest fuels and effects of moisture and wind [J]. International Journal of Wildland Fire, 2016, 26(1): 71–81.

[113] 李梦媛. 金属热颗粒引燃可燃堆垛材料的实验研究[D]. 中国科学技术大学, 2017.

[114] 李梦媛, 陈海翔, 张林鹤. 金属热颗粒引燃可燃堆垛材料的实验研究[J]. 火灾科学, 2017, 26 (3):140–146.

[115] Song J, Huang X, Liu N, et al. The wind effect on the transport and burning of firebrands[J]. Fire technology, 2017, 53(4): 1555–1568.

[116] Fang W, Peng Z, Chen H. Ignition of pine needle fuel bed by the coupled effects of a hot metal particle and thermal radiation[J]. Proceedings of the Combustion Institute, 2021, 38(3): 5101-5108.

[117] Manzello S L, Cleary T G, Shields J R, et al. On the ignition of fuel beds by firebrands [J]. Fire and Materials, 2006, 30: 77–87.

[118] Rowntree G W G, Stokes A D. Fire ignition by aluminium particles of controlled size [J]. Journal of electrical and electronics engineering, 1994, 14: 117–123.

[119] Yin P, Liu N, Chen H, et al. New correlation between ignition time and moisture content for pine needles attacked by firebrands [J]. Fire Technology, 2012, 50: 1–13.

[120] Urban J L, Zak C D, Fernandez-Pello C. Ignition behavior of powdered cellulose by hot steel spheres [C]. 8th US National Combustion Meeting, Canyons resort in Park City, Utah, 2013.

[121] Glushkov D O, Kuznetsov G V, Strizhak P A. Experimental and numerical study of coal dust ignition by a hot particle [J]. Applied Thermal Engineering, 2018, 133: 774–784.

[122] Glushkov D O, Legros J C, Strizhak P A, et al. Experimental and numerical study of heat transfer and oxidation reaction during ignition of diesel fuel by a hot particle [J]. Fuel, 2016, 175:105–115.

[123] Rallis C, Mangaya B. Ignition of veld grass by hot aluminium particles ejected from clashing overhead transmission lines [J]. Fire Technology, 2002, 38 (1): 81–92.

[124] Hadden R M, Scott S, Lautenberger C, et al. Ignition of combustible fuel beds by hot particles: An experimental and theoretical study [J]. Fire Technology, 2011, 47 (2): 341–355.

[125] Soulinaris G K, Halevidis C D, Polykrati A D, et al. Evaluation of the thermal stresses and dielectric phenomena in the investigation of the causes of wildfire involving distribution power lines [J]. Electric Power Systems Research, 2014, 117: 76–83.

[126] Zak C D. The effect of particle properties on hot particle spot fire ignition [D]. University of California, Berkeley, 2015.

[127] Urban JL, Zak CD, Fernandez-Pello C. Cellulose spot fire ignition by hot metal particles [J]. Proceedings of the Combustion Institute, 2015, 35 (3): 2707–2714.

[128] Urban J L, Zak C D, Song J, et al. Smoldering spot ignition of natural fuels by a hot metal particle [J]. Proceedings of the Combustion Institute, 2017, 36(2):3211–3218.

[129] Zak C D, Urban J L, Fernandez-Pello C. Characterizing the flaming ignition of cellulose fuel beds by hot steel spheres [J]. Combustion Science and Technology, 2014, 186 (10-11): 1618–1631.

[130] Fernandez-Pello C. Wildland fire spot ignition by sparks and firebrands [J]. Fire Safety Journal, 2017, 91:2–10.

[131] Urban J L, Song J, Santamaria S, et al. Ignition of a spot smolder in a moist fuel bed by a firebrand [J]. Fire Safety Journal, 2019, 108:102833.

[132] Urban J L, Zak C D, Fernandez-Pello C. Spot fire ignition of natural fuels by hot aluminum particles [J]. Fire Technology, 2018.

[133] Uhrlandt D, Gorchakov S, Brueser V, et al. Interaction of a free burning arc with regenerative protective layers [C]. 13th High-Tech Plasma Processes Conference, Journal of Physics: Conference Series, 2014, 550: 012010.

[134] Franke S, Methling R, Uhrlandt D, et al. Temperature determination in copper-dominated free-burning arcs [J]. Journal of Physics D: Applied Physics, 2014, 47 (1): 015202.

[135] Takaki A. On the Effect of Thermal Histories upon the Metallographic Structure of Electric Wire [J]. Research on Forensic Science, 1971, 24 (2): 48–56.

[136] 邓志谦. 铜及铜合金物理冶金基础[M]. 中南大学出版社, 2011.

[137] She H, Shu D, Dong A, et al. Relationship of particle stimulated nucleation, recrystallization and mechanical properties responding to Fe and Si contents in hot-extruded 7055 aluminum alloys [J]. Journal of Materials Science and Technology, 2019, 35:2570–2581.

[138] Wang H, Zhou C, Perry TA, et al. Effect of processing conditions and interfacial geometry on reaction metallurgical joining of copper [J]. Procedia Manufacturing, 2018, 26:1421–1428.

[139] Zhang J, Deng L. Metallographic microcosmic analysis on primary short circuited melted bead of copper wire heated in different temperature [J]. Procedia Engineering, 2013, 52: 583–587.

[140] Li Y, Liu X. Study on metallographic structure of melted breakpoint mark for copper wire current overloading [J]. Procedia Engineering, 2016, 135:482–485.

[141] Yang W, Ling S, Mo S, et al. Research of trace analysis and investigation method for refrigerator-caused fire [J]. Procedia Engineering, 2013, 52:526–531.

[142] Babrauskas V. Ignition Handbook [M]. Fire Science Publishers/Society of Fire Protection Engineers, Issaquah, WA, 2003.

[143] Ettling BV. Electrical wiring in building fires [J]. Fire Technology, 1978, 14: 317–325.

[144] 雷鹏奎. 引燃承载面对电焊熔珠痕迹特征的影响研究[D]. 西安科技大学，2019.

[145] 高伟, 吴莹, 刘术军, 等. 采AES深度刻蚀研究不同氧气含量下短路熔痕的成分[J]. 光谱学与光谱分析, 2010, 30 (7): 1999–2004.

[146] 吴莹, 孟庆山, 王新明, 等. 铜导线短路熔痕XPS研究[J]. 光谱学与光谱分析, 2010, 30 (5): 1408–1412.

[147] 王海蓉, 刘建勇, 姚浩伟, 等. 金属火灾熔痕的显微组织及物相成分的实验研究[J]. 光谱学与光谱分析, 2012, 32 (7): 1984–1988.

[148] 李阳, 何江涛. 基于Bayes判别模型的火场中铜导线短路熔痕定量金相鉴别方法研究[J]. 火灾科学, 2015, 24 (4): 201–208.

[149] 莫善军, 彭敬文, 梁栋. 电气火灾一次短路熔痕金相组织特征参数定量分析[J]. 中国安全生产科学技术, 2012, 8 (1): 63–70.

[150] 叶海伦, 梁栋, 林基深, 等. 不同燃烧环境下铝导线二次短路熔珠金相孔洞特征[J]. 消防科学与技术, 2017, 36 (3): 410–414.

[151] 姜蓬. 基于金相分析与烟熏图痕数值重构的火灾调查研究[D]. 合肥:中国科学技术大学, 2009.

[152] 张明, 邸曼, 夏大维. 铜导线短路熔痕内部孔洞形态特征参数的研究[J]. 消防科学与技术, 2011, 30 (7): 651–654.

[153] 王莉, 方仕童, 余圣辉, 等. 电气火灾中的电弧熔痕特征与外热关系[J]. 中山大学学报自然科学版, 2015, 54 (2): 1–7.

[154] Ao G, Chang-Zheng Z, Gang P, et al. Application of atomic force microscopy to assess a copper molten mark formed by short circuit [C]. 2014 Fifth International Conference on Intelligent Systems Design and Engineering Applications. IEEE, 2014: 222–225.

[155] Babrauskas V. Fire due to electrical arcing: can ‘cause’ beads be distinguished from ‘victim’ beads by physical or chemical testing? [J]. Fire and Materials, 2003: 189–201.

[156] Wright S A, Loud J D, Blanchard R A. Globules and beads: what do they indicate about small-diameter copper conductors that have been through a fire? [J]. Fire Technology, 2015, 51 (5): 1051–1070.

[157] Lee E P, Ohtani H, Matsubara Y, et al. Study on discrimination between primary and secondary molten marks using carbonized residue [J]. Fire Safety Journal, 2002, 37 (4): 353–368.

[158] Imada S, Hasegawa H. A study on discrimination between primary and secondary molten marks on blades of electric plugs by DAS [J]. Japanese Journal of Science and Technology for Identification, 2005, 10 (2): 147–156.

[159] 司永轩, 刘玲, 金南江, 等. 中美过电流火灾物证鉴定技术研究综述[J]. 消防科学与技术, 2017, 37 (3):419–422.

[160] Liu K H, Shih Y H, Chen G J. Microstructural study on oxygen permeated arc beads [J]. Journal of Nanomaterials, 2015, 6: 1–8.

[161] Liu K H, Shih Y H, Chen G J. Microstructural study on molten marks of fire-causing copper wires [J]. Materials, 2015, 8: 3776–3790.

[162] 徐学岩. 铝合金导线过电流故障演化过程及熔痕微观组织研究[D]. 西安科技大学.

[163] Babrauskas V. Arc beads from fires: can ‘cause’ beads be distinguished from ‘victim’ beads by physical or chemical testing? [J]. Journal of Fire Protection Engineering, 2016.

[164] Shaw C E. Fire marshals on duty [J]. NFPA J. 1965, 59: 26–2.

[165] Levinson D W. Copper metallurgy as a diagnostic tool for analysis of the origin of building fires [J]. Fire Technology 1977, 13: 211–222.

[166] Singh R P. Scanning electron microscopy of burnt electric wires [J]. Scanning Microscopy, 1987, 1 (4): 1539–1544.

[167] Erlandsson R , Strand G . An investigation of physical characteristics indicating primary or secondary electrical damage [J]. Fire Safety Journal, 1985, 8 (2):97–103.

[168] Ishibashi Y, Kishida J. Research on first and second fused mark discrimination of electric wires [J]. 1990 Annual Mtg. Japan Assn. for Fire Science and Engrg, 1990: 83–90.

[169] 于然, 付周兴, 王清亮, 等. 基于MATLAB的电弧建模仿真及故障分析[J]．高压电器，2011，47 (9): 95–99．

[170] 陈搏, 庄劲武, 肖翼洋, 等．基于半经验模型的高压限流熔断器电弧特性的计算与分析[J]．高电压技术，2014, 40 (1): 282–287．

[171] Cho D W, Song W H, Cho M H. Analysis of submerged arc welding process by three-dimensional computational fluid dynamic simulations [J]. Journal of Materials Processing Technology, 2013, 213 (12): 2278–2291.

[172] 孟宣宣, 王春明, 胡席远. 光纤激光焊接熔池和小孔的高速摄像与分析[J]. 电焊机, 2010, 40 (11): 78–81．

[173] 刘占民, 李明利, 薛龙. 熔化极电弧焊熔滴过渡过程的高速摄像[J]. 激光与红外, 2006, 36 (2): 131–134．

[174] 宋海鹰, 彭小齐, 刘征. 基于非接触式温度测量中的高温熔体识别方法[J]. 中南大学学报（自然科学版）, 2005, 36 (3): 426–430．

[175] 胡建国, 刘义祥, 覃萍. 火场中铝导线熔痕的微观形貌分析[J]. 理化检验-物理分册, 2004, 41 (8): 391–396.

[176] 胡建国, 田罡, 李阳. 火场中新型铜包铝导线熔痕的金相鉴别研究[J]. 消防科学与技术, 2014, 33 (11): 1347–1350.

[177] 李阳, 何江涛. 基于Bayes判别模型的火场中铜导线短路熔痕定量金相鉴定方法研究[J]. 火灾科学, 2015, 24(4): 201–208.

[178] 王元. 铜导线短路喷溅熔珠氧化物识别技术研究[D]. 中国人民武装警察部队学院，2016.

[179] Deng J, Li Y, Lü HF, et al. Metallurgical analysis of the ‘cause’ arc beads pattern characteristics under different short-circuit currents. Journal of Loss Prevention in the Process Industries, 2020, 68: 104339.

[180] Lü HF, Deng J, Li Y, et al. Influence of thermal environment on metallographic structure characteristics of the electric arc bead pattern [J]. Journal of Loss Prevention in the Process Industries, 2021, 70: 104426

[181] 中国国家标准化管理委员会. GB/T 16840-2012,电气火灾原因技术鉴定方法[S]. 北京:中国标准出版社, 2012.

[182] Gray D A, Drysdale D D, Lewis F. Identification of electrical sources of ignition in fires [J]. Fire Safety Journal, 1983, 6 (2):147–150.

[183] 王芸, 安晓利, 魏星, 等.铜导线短路熔痕微观形貌特征分析与研究[J]. 火灾科学, 2008 (01): 44–48.

[184] 王芸, 刘一乐, 舒中骏, 等. 铜导线二次短路迸溅熔珠密度特征实验研究[J]. 消防科学与技术, 2015, 34 (7): 970–974.

[185] 张臻. 铜导线一次短路线端熔珠与喷溅熔珠金相组织对比分析[J]. 分析仪器, 2016, 2: 77–80.

[186] 李九霖, 李阳, 刘义祥,等. 火场中铝合金导线熔痕的定量金相分析[J]. 消防科学与技术, 2020, 39 (4):565.

[187] 姜银松. 角接接头旋转电弧焊接数值模拟研究[D]. 南昌大学, 2016.

[188] 吴思根. 铝合金低频振荡扫描激光焊接数值模拟与试验研究[D]. 湖南大学.

[189] 汤昊. 基于Matlab高速影像识别技术的火灾前短路熔痕凝固过程研究[D]. 中国人民武装警察部队学院, 2018.

[190] 陈立辉. 基于 FLUENT 的火灾前短路熔痕组织形成研究[D]. 中国人民警察大学, 2019.

[191] 荣彦超, 李阳, 刘义祥, 等. 基于Matlab的短路熔痕凝固过程中熔体温度测算方法[J]. 中国安全生产科学技术, 2020, 16(8):7.

[192] Kong S G, Jin D, Li S, et al. Fast fire flame detection in surveillance video using logistic regression and temporal smoothing [J]. Fire Safety Journal, 2016, 79: 37–43.

[193] 李涛，向涛，黄仁杰，等．基于新的运动特征的火焰检测方法[J]．计算机仿真，2014，31(9): 392–396．

[194] 陈宗淇. 胶体与界面化学[M]. 高等教育出版社, 2001.

[195] 吴细秀. 开关电器触头材料喷溅侵蚀模型研究及其试验 [D]. 华中科技大学, 2005.

[196] Icove D J, Haynes G A. Kirk’s fire investigation [M]. Eighth Edition, Pearson Education, 2018.

[197] Zak C, Urban J, Tran V I, et al. Flaming ignition behavior of hot steel and aluminum spheres landing in cellulose fuel beds [J]. Fire Safety Science, 2014, 11: 1368–1378.

[198] Cook N R, Ridker P M. Advances in measuring the effect of individual predictors of cardiovascular risk: the role of reclassification measures [J]. Annals of internal medicine, 2009, 150 (11): 795–802.

[199] Bergman T L, Incropera F P, DeWitt D P, et al. Fundamentals of heat and mass transfer [M]. John Wiley & Sons, 2011.

[200] 彭志红. 飞火和热辐射点燃松针的实验及数值模拟研究[D]. 中国科学技术大学.

[201] Shi X, Zhang Y, Chen X, et al. Effects of thermal boundary conditions on spontaneous combustion of coal under temperature-programmed conditions [J]. Fuel, 2021, 295: 120591.

[202] 陈宗民. 铸造金属凝固原理[M]. 北京大学出版社, 2013.

[203] 张以忠.紫铜熔铸中如何防止吸气[J].上海有色金属，2005，(03):125–127.

[204] 常国威, 王建中. 金属凝固过程中的晶体生长与控制[M]. 冶金工业出版社, 2002.

[205] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016.

[206] Vapnik V. The nature of statistical learning theory [M]. Springer science & business media, 1999.

[207] Cortes C, Vapnik V. Support vector machine [J]. Machine Learning, 1995, 20:273–297.

[208] Floyd S, Warmuth M. Sample compression, learnability, and the Vapnik-Chervonenkis dimension [J]. Machine Learning, 1995, 21:269–304.

[209] Liu J, Bai M, Jiang N, et al. Structural risk minimization of rough set-based classifier [J]. Soft Computing, 2019:1–18.

[210] Kennedy J, Eberhart R. Particle swarm optimization [C]. Proceedings of the IEEE International Conference on Neural Networks, 1995.

[211] 曾建潮，介婧，崔志华. 微粒群算法[M].北京：科学出版社，2004.

[212] 李宁，邹彤. 基于粒子群的多目标优化算法[J].计算机工程与应用，2005，41(23)：81–83.

[213] 武宁. 基于SA-PSO的预拌混凝土配送车辆调度优化研究[D]. 河北工程大学, 2011.

[214] Rosenberg R S. Stimulation of genetic populations with biochemical properties: I. the model [J]. Mathematical Biosciences, 1970, 7(3-4): 223–257.

[215] Blickle T, Thiele L. A comparison of selection schemes used in genetic algorithms. TIK-Report 11, TIK Institut fur Technische Informatik und Kommunikationsnetze [J]. Computer Engineering and Networks Laboratory, ETH, Swiss Federal Institute of Technology, Gloriastrasse, 1995, 35(8092): 279–284.

[216] Abdi H, Williams L J. Principal component analysis [J]. Wiley Interdisciplinary Reviews Computational Statistics, 2010, 2 (4): 433–459．

[217] Pan S J, Yang Q. A survey on transfer learning [J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22 (10):1345–1359.

[218] 龙明盛. 迁移学习问题与方法研究[D]. 清华大学.2014.

[219] Pan S J, Tsang I W, Kwok J T, et al. Domain adaptation via transfer component analysis [J]. IEEE transactions on neural networks, 2010, 22 (2): 199–210.