摘要
由人工填土层和原始海相沉积土层组成的双层软土地基不同于常规地基形式,其沉降预测结果离散性大。天津滨海地区某工程地基为双层软土地基,采用真空预压法进行地基处理,通过现场监测揭示了双层软土地基沉降的特性。在沉降预测方面,利用卷积神经网络建立了双层软土地基沉降预测模型,并与双曲线拟合法进行比较,研究结果显示,该方法同样可以准确预测地基沉降变形,满足工程建设需求。
The double-layer soft soil foundation composed of artificial fill layer and original marine sedimentary soil layer is different from the conventional foundation form,and its settlement prediction results have large discreteness.The foundation of a certain project in the coastal area of Tianjin is a double-layer soft soil foundation.The vacuum preloading method was used for foundation treatment.Based on this engineering example,the settlement characteristics of the double-layer soft soil foundation were revealed through on-site monitoring.In terms of settlement prediction,a double-layer soft soil foundation settlement prediction model was established using convolutional neural networks,and compared with hyperbolic fitting method.The research results showed that this method can also accurately predict foundation settlement deformation,meeting the needs of engineering construction.
作者
孙冲
刘印鹏
陈少青
程林
Sun Chong;Liu Yinpeng;Chen Shaoqing;Cheng Lin(Tianjin Water Conservancy Engineering Group Co.,Ltd.,Tianjin 300222,China;Tianjin Beiyang Survey and Design Institute of Water Transport&Water Conservancy Co.,Ltd.,Tianjin 300460,China)
出处
《岩土工程技术》
2024年第3期287-293,共7页
Geotechnical Engineering Technique
基金
天津市水利工程集团有限公司科技项目(2022-02)。
关键词
双层软土地基
真空预压
沉降特性
机器学习
卷积神经网络
double-layer soft soil foundation
vacuum preloading
settlement characteristics
machine learning
convolution neural network