摘要
提出基于COWA-RBF神经网络的高层建筑火灾安全评价模型。从防火能力、灭火能力、安全疏散能力、安全管理能力4个方面构建评价指标体系。利用COWA算子对专家决策结果重新排序,加权求得指标权重,以期降低极值带来的不利影响。考虑到评价指标非线性关系和评价过程的动态性,利用RBF神经网络模拟高层建筑火灾安全等级。将模型运用到某住宅建筑的火灾安全评价,认为该建筑火灾安全等级高,装修材料耐火等级、报警器布置和达标情况、疏散通道畅通性、消防设施熟练程度4个指标需重点关注。
A fire safety evaluation of high-rise buildings based on COWA-RBF neural network was proposed.The evaluation index system was constructed from four aspects:fire prevention ability,fire extinguishing ability,safety evacuation ability,and safety management ability.Then the COWA operator was used to reorder the results of expert decision-making,and the weights of the indicators were weighted to reduce the adverse effects brought about by extreme values.Considering the nonlinear relationship between the evaluation index and the dynamics of the evaluation process,the RBF neural network is used to simulate the fire safety level of high-rise buildings.Finally,the model was applied to the fire safety evaluation of a residential building.The fire safety rating of the building was considered to be high,and the fireproof rating of the decoration materials,alarm arrangement and compliance,the patency of the evacuation channel,and the proficiency of the firefighting facilities were four key indicators.
作者
沈存莉
SHEN Cun-li(Chongqing Industry&Trade Polytechnic,Chongqing 408000,China)
出处
《消防科学与技术》
CAS
北大核心
2018年第10期1419-1422,共4页
Fire Science and Technology
作者简介
沈存莉(1984-),女,重庆工贸职业技术学院讲师,硕士,主要从事工程造价管理方面的研究,重庆市蒿枝坝都市工业园(涪陵区涪南路108号),408000。
