基于BP神经网络的城市区域火灾风险评估

Urban regional fire risk assessment with the method of BP neural network

针对火灾具有突发性、复杂性和不确定性的特点,提出了一种基于BP神经网络的城市区域火灾风险评估方法.从分析火灾发生的影响因素入手,建立了城市区域火灾风险评估BP神经网络模型,并利用层次分析法和专家调查法取得的样本进行网络训练和测试,取得良好效果,表明BP神经网络应用于火灾风险评估具有可行性和有效性.

In allusion to the actually that fire has the features of suddenness,complicacy and uncertainty,this paper gave out a method of urban regional fire risk assessment based on the BP network.,building a BP model of fire risk assessment from effect factors of fire,which was trained and tested with the samples gained by the methods of Analytic Hierarchy Process and Delphi method,the result shown that this model was feasible and effective to the fire risk assessment.