Early-warning and inducement analysis of super-large deformation of deep foundation pit on soft soil

In view of the large deformation in deep foundation pit engineering of a subway station in Nanjing soft soil area, firstly, the inducement of super-large deformation was analyzed based on the excavation revealing phenomenon and the trend of monitoring information. The inducement sensitivity was analyzed by using the grey relational analysis method. Secondly, according to the measured data, a correlation parameter weakening model was established to carry out reverse analysis to verify the cause of large deformation. Finally, based on the parameter inversion results, a weakening model system was constructed to evaluate the effectiveness of the deformation control measurement and predict the late deformation of the foundation pit. The prediction accuracy was verified by the ABC—BP neural network method. The results show that the adverse factors, such as pit side load, retaining pile stiffness weakening, poor soil quality and continuous precipitation, will cause the super-large deformation of deep foundation pit in soft soil. By parameter inversion, a parametric weakening model is obtained, which shows that the stiffness of pile is 82% of the design stiffness, the elastic modulus of soft soil is 35% of the geological survey value and the pile load at the pit side is 50 kN/m. The numerical analysis results are in good agreement with the monitoring results. The parameter weakening model can accurately predict the excavation deformation in the unexcavated area, and the result is similar to that of ABC—BP neural network. © 2022 Central South University of Technology. All rights reserved.