多数据融合的电气火灾危险性评估方法  

Electrical fire risk assessment method based on multi-data fusion

作　　者：颜峻 YAN Jun(Institute of Occupational Safety and Health, China Instituteof Industrial Relations, Beijing 100048, China)

机构地区：[1]中国劳动关系学院安全与职业卫生工程研究所

出　　处：《消防科学与技术》2019年第4期566-569,共4页Fire Science and Technology

基　　金：中国劳动关系研究生教育教学改革项目(YJG1702)

摘　　要：通过研究建筑电气系统运行状况参数间的逻辑关系,建立了基于贝叶斯信念网络的电气火灾危险性评估模型,实现了多数据融合下的电气火灾危险性评估与预警。评估方法将建筑供配电系统分为变配电装置、低压配电线路、照明装置和低压用电设备、接地和等电位联结等4个评估子单元,电气火灾危险性指数由4个子系统评估结果综合研判得到。方法基于逻辑判断对各子单元运行参数之间的关联和初始状态值进行了设定。通过实时监测电气系统运行参数的状态值,可以实现对电气火灾危险性的动态评估。By studying the logical relationship between parameters of building electrical system, electrical fire risk assessment model based on Bayesian belief network is established, which is used to assess the electric fire risk and early warning. The evaluation method divides the building power supply and distribution system into four sub- units, including transformer and distribution device, low voltage distribution line, lighting device and low voltage electrical equipment, grounding and equipotential connection. The electrical fire hazard index is comprehensively evaluated by the results of the 4 subsystems. The correlation and initial state values of the running parameters of each subunit were set based on logical judgment. The state values of the electrical system operation parameters are monitored in real time, and dynamic evaluation of electrical fire hazard can be done.

关 键 词：电气火灾 风险评估 贝叶斯信念网络 数据融合 

分 类 号：X913.4[环境科学与工程—安全科学] X934