摘要
基于目前在基坑支护结构中应用广泛的变形预测方法,有限元法和BP神经网络,结合基坑工程特点,提出将有限元模型与BP神经网络相结合的基坑变形预测方法.以某实际建筑基坑为研究对象,利用有限元软件进行近似建模,使用MATLAB软件实现有限元模型与BP神经网络模型相结合的预测模型,对基坑地下连续墙水平位移值进行预测对比.结果表明,有限元-BP神经网络预测模型预测值与实测值最为吻合,预测精度优于单一的有限元或BP神经网络预测模型.
Based on the widely used deformation prediction method,finite element method and BP neural network,combined with the characteristics of foundation pit engineering,a method to predict the deformation of foundation pit which combines the finite element model and BP neural network is put forward.Based on the foundation pit of a real building as the research object,using the finite element software to approximate modeling,using MATLAB software to realize the prediction model combining the finite element model and BP neural network model,the horizontal displacement values of the diaphragm wall of the foundation pit are predicted and contrasted.The results show that the predicted values of the finite element-BP neural network are the most consistent with the measured values,and the prediction accuracy is better than the single finite element or BP neural network prediction model.
作者
贾哲
郭庆军
郝倩雯
JIA Zhe;GUO Qingjun;HAO Qianwen(School of Civil and Architecture Engineering,Xi’an Technological University,Xi’an 710021,China)
出处
《河南科学》
2018年第3期430-435,共6页
Henan Science
基金
国家自然科学基金(51374165)
教育部人文社会科学研究(17YJC630032)
关键词
有限元
BP神经网络
地下连续墙
水平位移
监测数据
finite element
BP neural network
diaphragm wall
horizontal displacement
monitoring data
