Deformation Analysis of Existing Buildings Affected by Shield Tunnels Based on Intelligent Inversion and Measured Data

In the construction of urban underground shield tunnels, uneven deformation can easily occur when the shield passes through soft soil and other poor strata. Such deformation has a significant impact on surface settlement and may cause potential safety hazards to the surrounding existing buildings, directly affecting the safety of urban operation. When simulating and predicting surface settlements, the small-strain soil hardening model can more accurately characterize the mechanical parameters of soil. Nevertheless, its parameters are numerous and complicated to determine accurately, so parameter inversion is needed to determine the accurate parameters of the soft soil layer in order to more accurately predict the surface settlement. This study uses the EFAST method to analyse the sensitivity of the HSS model parameters of soft soil strata. It is determined that the parameters that have the most significant impact on the surface settlement are the reference tangent modulus, rebound modulus, and effective cohesion. Then, XGBoost’s fast calculation speed and high precision of SSA inversion are used to inverse and optimize the parameters with high sensitivity. Finally, according to the parameters of the soft soil layer obtained from inversion and measured data, the settlement deformation and safety behaviour of existing buildings are analysed. Combined with the actual shield tunnel project in a city along a river, the inversion calculation shows that the overall average error of the transverse monitoring section is 1.04 mm, and the average maximum error of each monitoring point in the overall shield process is 2.87 mm. The prediction effect is significantly improved compared with the original parameters. The accuracy of the inversion of soil layer parameters is verified from the perspective of time and space. The average settlement of the river embankment foundation is 2.5 mm. Compared with the original parameter data, the prediction results have been greatly improved, and the settlement deformation results are more consistent with the measured data.

Cette oeuvre a été publiée sous la license Creative Commons Attribution 4.0 (CC-BY 4.0). Il est autorisé de partager et adapter l'oeuvre tant que l'auteur est crédité et la license est indiquée (avec le lien ci-dessus). Vous devez aussi indiquer si des changements on été fait vis-à-vis de l'original.