灰关联与人工神经网络在建筑物震害预测中的应用

Forecasting seismic damage to buildings based on grey relation and artificial neural network model

基于灰关联识别方法,解析了各震害影响因子对多层砖房抗震性能的影响程度;并利用BP人工神经网络非线性模型对震害实例样本进行了训练。结果表明:利用灰关联分析,可得出各因子对多层砖房抗震性能影响程度的大小排序,有利于实际的工程抗震设计;基于BP人工神经网络模型的多层砖房的震害预测结果与震害实例的实际情况比较吻合,其思路和方法可推广于其他不同类型的建筑结构的震害预测。

The influence factors that affect the earthquake-resistance performance of the muhistory masonry buildings are analyzed. And the actual damage samples are trained by BP artificial neural network model. The research shows that the effect extent to the earthquake-resistance performance of these factors can be obtained by grey relation analysis and it can benefit the actual engineering design of earthquake-resistance and the prediction results are similar to the actual seismic and the analytic method the other structures with damage of multistory masonry buildings by the grey relation and BP neural network mode, and process this paper discussed can also be applied to the seismic damage prediction of different forms.