基于BP神经网络模型的郑州市城中村建筑震害预测  

Earthquake Damage Prediction of Buildings in Zhengzhou Urban-villages Based on BP Neural Network Model

作　　者：王建强[1] 陈志军 田野[1] 赵军[1] 

机构地区：[1]郑州大学土木工程学院,河南郑州450001 [2]珠海市规划设计研究院,广东珠海51900 [3]河南省建筑科学研究院有限公司

出　　处：《工程抗震与加固改造》2015年第5期135-140,共6页Earthquake Resistant Engineering and Retrofitting

基　　金：河南省重点科技攻关计划项目(132102310447);河南省高校科技创新团队(15IRTSTHN026)

摘　　要：通过对郑州市两个城中村建筑基本情况的调研,确定了影响房屋震害的主要影响因素,对典型房屋进行了抗震性能分析,进而建立了城中村建筑的震害样本库和BP神经网络模型,对城中村建筑的震害进行了预测,得到了城中村建筑的震害矩阵,结果表明,城中村建筑抗震性能较差,7度多遇地震作用下绝大部分建筑将产生中等以上的破坏,将造成重大的人员伤亡和财产损失。Through the survey of the buildings in Zhenzhou urban-villages, the characteristics of the buildings are determined, the seismic performance of the typical buildings is analyzed. The earthquake damage sample library and BP network model are established. The earthquake damage prediction of buildings in Urban-villages is carried out, so the damage matrix of buildings is obtained. The results show that the seismic performance of the buildings are very poor, most of the buildings will be destroyed in frequent earthquake, and it will cause significant casualties and property losses.

关 键 词：城中村 抗震性能 BP神经网络 震害预测 

分 类 号：P315.9[天文地球—地震学] TU352.11[天文地球—固体地球物理学]