主管:中华人民共和国应急管理部
主办:应急管理部天津消防研究所
ISSN 1009-0029  CN 12-1311/TU

消防科学与技术 ›› 2022, Vol. 41 ›› Issue (7): 942-945.

• • 上一篇    下一篇

基于未确知聚类的高层建筑防火风险评估系统

安 晶,张立宁,庞晓娜,崔彩云   

  1. (华北科技学院,河北 廊坊065201)
  • 出版日期:2022-07-15 发布日期:2022-07-15
  • 作者简介:作者简介:安 晶(1981-),女,辽宁大连人,华北科技学院电信学院讲师,硕士,主要从事智能算法与信息处理方面的研究,河北省三河市燕郊开发区学院街467号,065201。
  • 基金资助:
    国家自然科学基金项目(72001079);河北省国际合作基地项目(20594501D);河北省教育厅科研项目(Z2020117);中央高校基本科研业务费资助项目(3142018070,3142014075)

The fire risk assessment system for high-risebuildings based on unascertainedclustering method

AN Jing, ZHANG Li-ning,PANG Xiao-na, CUI Cai-yun   

  1. (North China Institute of Science and Technology, Hebei Langfang 065201, China)
  • Online:2022-07-15 Published:2022-07-15

摘要: 目前,高层建筑的火灾防控难题一直未得到有效解决。针对高层民用建筑防火能力风险评估中评估指标获取不科学、指标体系差异大的问题,运用典型火灾事件分析法,结合《建筑设计防火规范》,科学获取影响高层建筑防火能力的12个主要不确定性风险因素,构建了更为科学的高层建筑防火能力风险评估指标体系。在此基础上,引入未确知聚类理论,构建了基于未确知聚类的高层建筑防火能力风险评估模型。并以8栋典型高层民用建筑火灾风险数据为例,进行实证分析。研究成果表明,该风险评估系统可行有效。研究成果可为高层民用建筑火灾风险防控提供一种新的方法和途径。

关键词: 关键词:高层民用建筑, 典型事件分析, 未确知聚类, 防火风险评估

Abstract: Abstract: The fire prevention and control problem of high-rise buildings has not been effectively solved. Aiming at the problems of the evaluation indicators acquisition that is unscientific and the index system varies greatly in high-rise civil building fire risk assessment, the typical fire event analysis method was used in this study, and combining with the code for fire protection design of building, 12 major uncertainty risk factors affecting the fire prevention capacity of high-rise buildings were obtained, to build a more scientific fire risk assessment index system. Then the unascertained clustering theory is introduced to build a new high-rise building fire risk assessment model. And the fire risk data of 8 typical high-rise civil buildings were taken as examples to conduct empirical analysis. Study shows that the system is effective. The research results can provide a new way for the fire prevention and control of high-rise civil buildings.

Key words: Key words: high-rise civil building, the typical event analysis;the unascertained clustering model, fire risk assessment