铜导线短路熔痕的SEM/EDS分析

在电气线路火灾原因的调查中,为更好地区分铜导线一次短路与二次短路的熔痕特征,用扫描电子显微镜(SEM)和X-射线能谱仪(EDS)对2种熔痕的形貌和相应的元素成分进行了试验研究。结果表明:铜导线一次短路与二次短路熔痕的表观形貌特征和元素成分具有明显的不同,二者的区别主要表现在熔痕表面所含的气孔、元素种类及导线本体与熔痕之间的过渡区方面,一次短路熔痕表面气孔多,元素种类少,导线与熔痕之间有明显界限,二次短路熔痕则相反。该研究结果可以区分短路熔痕形成的环境气氛,以此判断导线短路熔痕类型,为电气线路火灾原因的调查提供科学依据。     

基于PCA-BP融合的边坡稳定性评价模型

针对边坡岩土体物理力学性质复杂、边坡稳定性影响因素众多等特点,提出将主成分分析(PCA)和BP神经网络结合起来进行边坡稳定性评价的方法。若BP神经网络训练误差一定,则网络信息容量与样本数成正比。当样本数较少时,就必须减少样本维数,以达到较好的匹配效果,为此,引入主成分分析法(PCA)对影响边坡稳定的众多变量进行降维处理,以消除输入数据间的相关性,有效地减少预测模型的输入量,优化网络的输入节点数,提高网络的运行效率。针对BP算法容易落入局部最小、收敛速度慢等缺点,引入粒子群优化算法(PSO)优化神经网络的连接权重与阀值,从而克服了BP神经网络的固有缺陷。在此基础上,建立基于PSO优化算法的PCA-BP融合的边坡稳定性评价模型。模型分为3个层次,第一层次为输入层,即经过PCA分析之后获得的主成分;第二层次为隐含层;第三层次为输出层,即安全系数。应用该评价模型进行算例分析,结果表明,安全系数的模型计算值与参考值的绝对误差均很小,相对误差均控制在6%以内,吻合程度较高。     

铜导线短路熔痕微观形貌特征分析与研究

采用模拟实验的方法制备了铜导线一次短路熔痕和二次短路熔痕,利用KYKY-2800B型扫描电子显微镜(SEM)对铜导线一次短路熔痕和二次短路熔痕的微观形貌进行了观察,并对内部气孔的数量、大小、分布等进行了统计分析,发现了一次短路气孔的数量、分布与二次短路之间的差别与长期采用的鉴定依据不同.     

消防射水对导线火烧熔痕金相组织的影响

模拟火灾现场,制作不同线径铜导线的火烧熔珠,然后在不同温度下对熔珠进行冷却.利用金相显微镜对火烧熔珠的金相组织进行观察,发现铜导线火烧熔痕在不同温度下经消防射水冷却后,会出现类似一次短路和二次短路熔痕的金相组织.     

基于PCA-ELM的地震死亡人数评估

地震往往会在短时间内造成巨大的人员伤亡和财产损失,为了震后救援工作能够快速高效的展开,建立了主成分分析(Principal Components Analysis,PCA)和极限学习机(Extreme Learning Machine,ELM)相结合的地震死亡人数预测模型。综合考虑影响地震死亡人数的多种因素,选取震级、震源深度、震中烈度、抗震设防烈度、人口密度、发生时间和预报水平7项主要因素作为评价指标。首先,利用主成分分析对原始变量进行降维处理,并计算出主成分得分,作为ELM的输入;其次,对构建的ELM地震死亡人数预测模型进行训练;最后,对选取的32组地震样本进行算例仿真,并与未经PCA处理的ELM算法和BP神经网络算法进行对比。结果表明,基于主成分分析的极限学习机算法对地震死亡人数预测具有较高的预测精度,验证效果良好。     

含硫气井完整性风险等级预测研究

为保证含硫气井在开发过程中安全高效生产,提出一种基于主成分分析(PCA)和BP神经网络的含硫气井完整性风险等级预测模型。首先,采用Bow-tie方法得到含硫气井完整性失效风险因素,通过模糊评价法对风险因素进行量化处理;然后,利用PCA提取综合指标,并结合BP神经网络得到预测模型,进而得到风险等级。研究结果表明:通过PCA能使BP神经网络的输入数据由28个减少至4个,所建模型的风险等级预测精度高于未经PCA的BP神经网络;通过PCA与BP神经网络结合的预测模型可识别含硫气井开发过程中完整性失效的风险因素,完善了含硫气井风险等级预测技术。     

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

借助计算机图像处理技术,改进了定量金相分析方法,测量了20个一次短路熔痕和20个二次短路熔痕的金相组织,对数据进行了主成分分析,提炼出了晶粒特征因子、气孔特征因子和复合因子三个主成分,基于Bayes判别法建立了火场中短路熔痕定量金相分析判别模型,经自身检验和交互检验,准确率均高于80%,充分满足鉴定要求,为更加有效地应用定量金相法鉴别短路熔痕提供了理论依据。     

电气火灾一次短路熔痕金相组织特征参数定量分析

电气火灾一次短路熔痕鉴别和判定对分析火灾原因和事故认定至关重要,而目前金相分析方法目前处于经验和半经验水平,缺乏定量分析手段.本文开展不同电流条件下(100A至300A,间隔为20A)铜导线一次短路模拟实验,制作相应短路熔痕金相图谱.同时引入描述金相组织特征参数,基于Image - ProPlus应用数字图像处理技术提取图谱的金相组织特征参数量化指标,讨论金相组织晶粒和孔洞尺寸变化的特点,实现金相微观组织特征参数和电流之间的关系数据对比分析,探讨一次短路熔痕金相组织的量化判据和变化规律.     

基于SKPCA与NEAT算法的煤与瓦斯突出危险性预测

针对煤与瓦斯突出预测精度不足的问题,为了减少突出灾害的危害,提出了基于稀疏核主成分分析法(SKPCA)与增强拓扑神经进化算法(NEAT)的煤与瓦斯突出危险性预测方法.通过核主成分分析法对突出非线性数据进行降维,然后对主成分数据进行稀疏化,减少原始数据中不重要元素对降维后主成分的影响.搭建增强拓扑神经进化网络(NEAT)预测模型,采用进化算法同时优化神经网络的权值和拓扑结构,得到最佳神经网络预测模型.将处理后的主成分数据输入模型预测突出危险指数,结果表明,该方法危险等级预测准确度达到98％,SKPCA-NEAT模型在煤与瓦斯突出预测上相比PCA和BP神经网络具有优势.     

基于Matlab的短路熔痕凝固过程中熔体温度测算方法*

为解决无法对瞬时凝固的熔痕高温熔体进行测温，从而致使火灾物证鉴定界对短路熔痕组织形成规律无法达成共识的问题，使用高速摄像机采集熔痕形成时熔体的彩色影像，运用Matlab图像色温识别技术，对熔体凝固时的色彩变化进行识别与分析，从而建立测算短路熔痕凝固速率的方法。结果表明:该方法的平均标准差为9.63，标准不确定度为2.55 K，计算结果满足测试要求；在测算的10个熔痕中，灭弧瞬间熔体温度为1 734.03~2 759.15 K，初始冷却速率为-255.16~-86.61 K/ms，凝固时长为46.45~87.99 ms；灭弧时熔体温度与初始冷却速率呈近似正相关关系，与熔痕体积无关。通过此方法测算短路发生时高温熔体的温度变化，可为研究短路故障迸溅熔痕引燃、熔痕组织变化规律提供方法支持。     

Numerical simulation and experimental study on metallographic structure characteristics of melting trace caused by overcurrent fault of copper wire

The microscopic characteristics of melting trace caused by copper wire fault are an important basis for the physical evidence identification of electrical fires. Metallography is an important method in electrical fire research, and computer-aided quantitative analysis on metallographic structure is common in materials aspect. In this study, the metallographic structure and phase composition characteristics of melting trace caused by overcurrent fault were explored with the aid of a metallographic microscope and an X-ray diffractometer (XRD). The results show that the grains are mainly slender dendritic and columnar crystals when the current is 128 A, while they are coarse dendritic and cellular crystals when the current is greater than 192 A; the average grain diameter of melting trace of copper wire grows with the increase in current. The main phases of melting trace caused by overcurrent fault of copper wire are α-Cu base and Cu₂O, and the new phases Cu₄Si and Cu₂Mg are formed when the current is higher than 160 A. In addition, the solidification process of metallographic structure of copper wire under different currents was simulated by the Procast software. It is found that the simulation results are basically in agreement with the experimental data, which suggests that this model can effectively predict the microstructure characteristics of melting trace of copper wire under different currents. The research results can provide a sound basis for the physical evidence identification of electrical fire and improve the accuracy and scientificity of electrical fire investigation.     

Metallurgical analysis of the ‘cause’ arc beads pattern characteristics under different short-circuit currents

Electric arc beads commonly cause large fire and explosion accidents, especially during the powder process industries. During such fires, the original equipment is often severely burned, rendering electrical fire investigation impossible. Therefore, identifying the beads caused by electrical faults through metallurgical investigation is crucial. This study focused on the ‘cause’ of arc beads and determined their pattern characteristics through metallurgical analysis of copper wires with short-circuit currents of 100, 160, 200, and 240 A. According to microstructure characteristics and metallographic knowledge, bead pattern characteristics were divided into six types: hypereutectic (A), dendritic eutectic (B), chilled layer (C), cellular crystal (D), dendritic (E), and coarse columnar crystal (F). Oxide levels of each category were A (4.5%–7.6%) > C (2.0%–3.8%) > B (1.1%–2.4%) > D (1.0%–2.1%) > E (0.6%–1.9%) > F (0.1%–0.7%). When the short-circuit currents were the same, there were a greater number of small particle beads than large ones. As the current increased, the number of distributed beads with a particle size d < 1 mm decreased, with 1 mm ≤ d < 2 mm increased, and with d ≥ 2 mm were stable. Bead size and distribution had a negative linear correlation at 200 A. When the current increased, the proportion of beads with A and C structures gradually decreased, and those with D structures increased to more than 50%. Oxygen content and arc temperature determined the oxide form. These results are vital for electrical fire investigations and provide theoretical support for fire material evidence extraction and cause identification.     

基于PCA-FPP-BP神经网络的高层建筑火灾安全评价

为解决高层建筑构造复杂、人员密度大、火灾触发因素繁多而造成高层建筑火灾安全评价困难的问题,本文提出基于PCA-FPP-BP神经网络的高层建筑火灾安全评价模型。首先运用主成分分析(PCA)对构建的高层建筑火灾安全评价指标降维处理,筛选主要信息;接着基于三角模糊数构建模糊评判矩阵,利用模糊优先规划(FPP)求解指标的权重值,减少主观的影响;最后考虑到指标间关系错综复杂彼此交叉和反馈的特性,选择BP神经网络对高层建筑火灾安全进行评价。通过工程案例证明该评价模型的实用性以及可靠性。     

电气火灾残留物的SEM分析方法研究

通过分析电气火灾的形成原因,总结出该类型火灾的残留物种类。然后利用电焊机和电阻炉等设备分别制备出铜铝导线的一、二次短路熔痕和火烧熔痕,电热丝的通电过热和火烧熔痕,并运用了目前一种新式有效的显微结构分析工具——扫描电子显微镜,对以上样品的表面进行微观形貌的观察、比较与分析。通过分析讨论,可以发现电气火灾残留物在火灾前后其表面的微观形貌特征存在明显不同。根据这些特征能够很轻易的鉴别出电气设备在火灾发生前的使用状态,从而推断出火灾是否由该电气设备引发而起。此项结论能够为从事火灾调查的人员对电气火灾原因的准确认定提供可靠依据,也能够为扫描电子显微镜技术的应用领域进行拓宽提供一些帮助。     

改进的GA-BP神经网络在矿井突水水源判别中的应用

矿井突水水源的判别是制定防治水措施的重要环节。通过对某矿含水层水化学特 性的相关性分析,将PCA算法、K折交叉验证算法嵌入GA-BP神经网络,提出了一种新的 GA-BP神经网络,将其应用于实例分析中,并与传统的方法进行比较。结果表明:针对 水化学特性相近的含水层,PCA算法能够排除样本中的冗余信息,降低样本指标维度, 简化BP神经网络结构;K折交叉验证算法能够提高GA算法对BP神经网络权值的寻优质量 ,使GA算法的进化方向更具合理性;二者的引入大大优化了传统GA-BP神经网络性能, 其判别精度更高、适用性更强、结果更可靠,在矿井突水水源判别方面具有很好的应用 前景。     

Influence of thermal environment on metallographic structure characteristics of the electric arc bead pattern

Electrical apparatuses are prone to arc, which generally causes a fire, even an explosion hazard, when a flammable gas mixture is present, especially during industrial processes. Terrible fire scenes are challenging for fire investigations. In this work, by performing a simultaneous thermal analysis test we simulated a fire environment and found that as the oxygen concentration decreased, the oxidation/exothermic peak temperature of ‘cause’ bead became higher, but the melting temperature was unaffected. Results indicated that the bead pattern underwent oxidation at approximately 831 °C, melting initiated at approximately 1060 °C, and the pattern then disappeared. The melted pattern grain changes were divided into three critical temperature stages: Approximately 600 °C, the onset temperature at which the melted pattern grains began to be equiaxed; approximately 831 °C, at which the grains were interspersed with oxygen-containing material; and 831–1060 °C, when the grains disappeared, which is a criterion for identifying electrical fires. However, the boundaries remained throughout the thermal environment process. Moreover, the bead pattern demonstrated three metallographic regions: Deep layer (Region I), the intermediate layer (Region Ⅱ), and surface layer (Region Ⅲ). Region I was the most thermally sensitive, in which equiaxed crystals first appeared. Region Ⅲ was the thermal reaction lag zone, in which the typical branching crystals finally disappeared, and Region Ⅱ was intermediate between Regions I and Ⅲ. The results may help fire investigators determine the fire scene temperature stages and provide support for fire evidence extraction.     

单芯铜线过电流故障电弧熔痕的微观特征研究*

为准确认定电气火灾中过电流故障，模拟4~7倍额定电流（Ie）条件下单芯铜导线过电流故障，利用光学显微镜和扫描电子显微镜（SEM）分析过电流故障电弧熔痕的组织特征，并结合Image-Pro Plus软件测定晶粒直径、周长和面积，探究金相组织的量化判据和变化规律。结果表明:当I＝4Ie时，过电流故障电弧熔痕的金相组织主要为方向性较强的树枝晶；当I＝5Ie，6Ie时，金相组织内树枝晶占比减小，胞状晶占比升高；当I＝7Ie时，方向性较弱的胞状晶占主导。随着电流增加，晶粒的平均直径、周长和面积均呈增长趋势。通过火灾案例分析可知，实验结果与现场的电弧熔痕组织特征基本相符，验证火灾事故原因为过电流故障。