淤泥质软土地层盾构施工地表变形预测  

Prediction of Surface Deformation during Shield Tunneling Construction in Soft Soil Layers with Silt

作　　者：张旭生 ZHANG Xusheng(Jiangsu Provincial Transportation Engineering Construction Bureau,Nanjing Jiangsu 210001 China)

机构地区：[1]江苏省交通工程建设局,江苏南京210001

出　　处：《江苏建筑》2025年第1期51-55,共5页Jiangsu Construction

摘　　要：由于淤泥质软土强度低、易扰动的特点,盾构施工穿越软土地层时容易引起地表大变形,对地面建筑物产生不良影响。文章针对盾构隧道施工穿越淤泥质软土地层导致地表变形超限的问题,提出了利用机器学习方法建立盾构施工地表变形智能预测模型。基于工程现场实测的数据,以盾构刀盘推力、刀盘扭矩、切口压力等参数作为输入矩阵,地表测点变形为输出向量构建机器学习预测模型,实现了开挖面前方地表变形的提前预测,为现场技术人员调整盾构参数、提前采取变形控制措施提供了参考。Due to the characteristics of low strength and easy disturbance of silty soft soil,shield tunnel construction could easily cause large deformation of the ground surface when crossing the soft soil layer,which will have a negative impact on ground buildings and structures.Aiming at the problem of surface deformation exceeding the limit caused by the shield tunnel construction crossing through the silt soft soil layer,this paper established an intelligent prediction model of the surface deformation caused by shield construction through the machine learning method.Based on field measurement data,a machine learning prediction model was constructed using parameters such as shield cutter thrust,cutter torque,and face pressure as the input matrix,with surface deformation measurements as the output vector.This model enables the early prediction of surface deformation ahead of the excavation face,providing a reference for on-site technicians to adjust shield parameters and implement deformation control measures in advance.

关 键 词：淤泥质软土 盾构施工 变形预测 机器学习 卷积神经网络 

分 类 号：TU43[建筑科学—岩土工程]