基于贝叶斯网络的大型城市社区消防安全评估模型*

针对大型城市社区的消防安全预控管理问题，分析了国内外消防安全评估的研究现状，提出了一个基于贝叶斯网络的消防安全评估模型，并结合层次分析法对消防安全进行定量评估。该模型利用贝叶斯网络推理反映评估指标之间的条件相关关系，运用层次分析法确定指标权重，有利于提高大型城市社区消防安全评估的客观性、准确性、全面性及智能性，进而改善城市社区消防安全预控管理的预测能力和服务效率，且实验结果证明了该模型的有效性和实用性。

Fire safety evaluation model of large city community based on Bayesian network

Aimed at the problem of fire precontrol and management in large city community, this paper analyzed the present research of fire safety evaluation inside and outside, put forward a fire safety evaluation model based on Bayesian network. The model could be used to evaluate the state of fire safety quantitatively combined with analytic hierarchy process(AHP). It reflected the causal relationship among evaluation indicators by Bayesian network inference more objectively and determined the weight of indicators by AHP more accurately. It has advantages of improving the objectivity, accuracy, comprehensive and intelligence of fire safety evaluation, and ameliorating the prediction capabilities and service efficiency of fire safety precontrol and management in large city community. At last, the experimental result shows the validity and practical of the model.