摘要
城市遭受地震袭击后伴随大火而引起房屋烧毁与财产损失是巨大的．这与地震烈度、火源发生与蔓延、灭火能力有关．文中利用人工智能神经网络方法预测震后火灾损失；根据美国北加州湾区和南加州洛杉矶地区的有关历史资料，训练种经网络对某些地震烈度的火灾损失作出预测．该方法能迅速、准确地预测未来指定烈度与气候条件下的火灾损失．对其它自然灾害，在一定条件下也可应用．
The heavy losses of burnt buildings and properties by fires following earthquakes are mainly related with quake intensity, fire ignition and its spread,and fire suppresion. In this paper, the fire losses under future earthquakes of specified intensity is predicted by using Al neural networks approach based on the historical information of Bay Area and L.A. California, USA.
出处
《同济大学学报（自然科学版）》
EI
CAS
CSCD
1995年第1期15-20,共6页
Journal of Tongji University:Natural Science
关键词
城市
地震
火灾
神经网络
预测
Fire following earthquake
Neural networks
B-P algorithm
Processing element
