基于灰色关联分析的变压器油池火灭火有效性评估*

为对比不同灭火系统扑灭变压器油池火灾有效性，提升变压器油池火灭火效能，从效能评估工具、灭火临界参数、行业法规中提取并确定变压器油池火灭火有效性评估指标集，包括3个观测变量，6个具体指标。采用Otsu’s算法处理火灾图像确定火焰面积，通过层次分析法确定各观测变量和具体指标权重，应用灰色关联法构建变压器油池灭火有效性评估模型。以水喷雾、细水雾、泡沫-水喷雾以及泡沫-细水雾扑灭变压器油池火为例，分别计算4种不同灭火系统下的灭火有效性。研究结果表明:灭火方案有效性为泡沫细水雾（1.000）>纯水细水雾（0.653）>泡沫水喷雾（0.447）>纯水水喷雾（0.333）。该评估模型可以将消防监测终端采集的消防大数据与人工智能技术有效结合，有助于提升变压器消防安全管理智能化水平。

Effectiveness evaluation on fire extinguishing of transformer pool fire based on grey correlation analysis

In order to compare the effectiveness of different fire-fighting systems in extinguishing the oil pool fire of transformers,and improve the fire extinguishing efficiency of transformer oil pool fire,the index set of the effectiveness evaluation on fire extinguishing of transformer pool fire was extracted from the effectiveness evaluation tools,fire extinguishing critical parameters and industrial regulations and determined,including 3 observation variables and 6 specific indexes.The Otsu’s algorithm was used to process the fire image to determine the flame area,then the weights of each observation variable and specific index were determined by analytic hierarchy process,and the grey correlation method was used to construct the effectiveness evaluation model for the extinguishing of transformer oil pool fire.Taking the water spray,water mist,foam-water spray and foam-water mist extinguishing transformer oil pool fire as examples,the fire extinguishing effectiveness under four different fire extinguishing systems was calculated respectively.The results showed that the effectiveness of the fire extinguishing scheme is foam water mist (1.000)>pure water mist (0.653)>foam water spray (0.447)>pure water spray (0.333).The evaluation model can effectively combine the fire big data collected by the fire monitoring terminals with the artificial intelligence technology,which is helpful to improve the intelligent level of fire safety management on the transformers.