摘要
以某地下商城开发项目为例,利用深基坑开挖期间周边建筑物沉降数据,分别采用灰色GM(1,1)模型、BP神经网络模型、灰色BP神经网络组合模型对沉降数据进行预测,将三种模型预测数据与实测数据通过图表的方式进行对比分析,得出灰色BP神经网络组合模型相对于两种单一模型预测精度较高、预测沉降值更准确的结论,进一步为及时采取预防措施和避免灾难的发生提供了可靠的科学依据。
This paper takes an underground shopping mall development project as an example to use the cumulative settlement of a building measured during the excavation of a deep foundation pit.The gray GM (1,1) model, BP neural network model, and gray BP neural network combination model are used for settlement.The data is used for forecasting, and the three types of model prediction data and actual measurement data are compared and analyzed through charts.It is concluded that the gray BP neural network combination model has higher prediction accuracy with respect to the two single types, and the predicted settlement value is more accurate, which can provide a reliable scientific basis for taking preventive measures in time and avoiding the occurrence of disasters.
作者
查天宇
成枢
吕磊
ZHA Tianyu;CHENG Shu;LYU Lei(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China)
出处
《测绘与空间地理信息》
2019年第9期212-215,共4页
Geomatics & Spatial Information Technology
关键词
深基坑
沉降预测
灰色模型
神经网络
deep foundation pit
settlement forecast
gray model
neural networks
