某深基坑施工变形监测与预测研究

Research on Deformation Monitoring and Prediction of a Deep Foundation Pit Construction

基坑开挖变形预测是主动风险控制过程,以烟台市某高层框架式剪力墙结构住宅楼项目基坑工程为研究对象,运用现场实测的方法获得基坑支护桩水平位移和地表沉降实测值,并基于BPNN与NSGA-Ⅱ组合算法对基坑变形进行预测,研究了基坑开挖过程中的支护桩及地表变形特征。研究结果表明,基于BPNN与NSGA-Ⅱ组合算法的基坑地表沉降、支护结构水平位移测试样本决定系数与训练样本决定系数均具有显著的相关性;在开挖段,支护桩水平位移均呈非线性增加趋势,而在嵌固段则呈近线性增加;随着时间的增加,地表沉降均呈现迅速下降并在60d后趋于收敛;BPNN与NSGA-Ⅱ组合算法具有较高的基坑变形预测精度。

The prediction of deformation during excavation of foundation pit is an active risk control process.Taking the foundation pit engineering of a high-rise frame shear wall structure residential building project in Yantai City as the research object,the method of on-site measurement was used to obtain the measured values of horizontal displacement and surface settlement of foundation pit support piles.Based on the combination algorithm of BPNN and NSGA-Ⅱ,the deformation of foundation pit was predicted,and the characteristics of support piles and surface deformation during excavation of foundation pit were studied.The research results indicate that there is a significant correlation between the determination coefficients of test samples and training samples for foundation pit surface settlement and horizontal displacement of support structures based on the combination algorithm of BPNN and NSGA-Ⅱ;In the excavation section,the horizontal displacement of the supporting piles shows a non-linear increasing trend,while in the embedded section,it shows a nearly linear increasing trend;With the increase of time,surface subsidence shows a rapid decline and tends to converge after 60days;The combination algorithm of BPNN and NSGA-Ⅱhas high accuracy in predicting foundation pitdeformation.