Abstract:Monitoring and predicting ground settlement during tunnel construction is of paramount importance for ensuring the safety of tunnel construction and the stability of the surrounding environment. Existing studies on settlement prediction mainly rely on single settlement values and often overlook temporal characteristics, and that prediction models struggle to capture the nonlinear trends in actual settlements, leading to suboptimal predictive accuracy. In this study, based on the monitoring data of settlement deformation in a subway section of a certain city, a multi-factor-driven prediction method for surface settlement in subway tunnel excavation using the Informer model is proposed. The predictive accuracy is compared and analyzed with other models, including CNN-LSTM, LSTM, SARIMA, and Transformer. The results indicate that: (1) The Informer model outperforms CNN-LSTM, LSTM, SARIMA, and Transformer in terms of RMSE, MAE, and MAPE evaluation metrics, demonstrating that the Informer model exhibits smaller average prediction errors in forecasting surface settlement during subway construction. (2) Compared to LSTM, CNN-LSTM, Transformer, and other models, the Informer model can better capture the temporal characteristics and long-term forecasting ability of settlement data, while the SARIMA model fails to capture the temporal features in actual settlement data effectively. (3) Considering the influencing factors of temperature and soil pressure has a positive impact on the predictive performance of the Informer model, and the relationship between soil pressure information in the case study construction area and surface settlement is more closely associated. In summary, the Informer model, which takes into account temporal characteristics and multiple influencing factors, demonstrates good predictive ability for nonlinear settlement data. It provides a new method for analyzing settlement trends caused by subway tunnel excavation under complex environmental conditions, facilitating efficient and accurate assessment. It also offers objective data support for short-term bridge construction scheduling and long-term construction planning.

Settlement Estimation During Foundation Excavation Using Pattern Analysis and Explainable AI Modeling

Support Vector Machine Model of Settlement Prediction of Road Soft Foundation

Spatial-temporal Fusion Network for Maximum Ground Surface Settlement Prediction During Tunnel Excavation

ANN-Based Dynamic Prediction of Daily Ground Settlement of Foundation Pit Considering Time-Dependent Influence Factors

Real-time online prediction of surface settlement considering spatiotemporal characteristics during foundation excavation

A novel back-analysis method for improving predictions of time-dependent ground settlements in deep excavations in soft soils

Fitting and prediction of observational settlement for soft ground

Research on Prediction Model of Support Vector Machine Based Land Subsidence Caused by Foundation Pit Dewatering

Multi-objective Optimization-Based Prediction of Excavation-Induced Tunnel Displacement

Prediction of surface settlement around subway foundation pits based on spatiotemporal characteristics and deep learning models

Transformer based neural network for daily ground settlement prediction of foundation pit considering spatial correlation

Predictions of Deep Excavation Responses Considering Model Uncertainty: Integrating BiLSTM Neural Networks with Bayesian Updating

Developing effective optimized machine learning approaches for settlement prediction of shallow foundation

Bayesian Inference of Spatially Varying Parameters in Soil Constitutive Models by Using Deformation Observation Data

A multi-factor-driven approach for predicting surface settlement caused by the construction of subway tunnels by undercutting method

Improved Machine Learning Model for Urban Tunnel Settlement Prediction Using Sparse Data

New Approach to Ground Settlement Analysis During Box Jacking in Spatially Variable Soils

Study on settlement prediction of soft ground considering multiple feature parameters based on ISSA-RF model

Application of GA-BP Neural Network Optimized by Grey Verhulst Model around Settlement Prediction of Foundation Pit

Settlement Research and Prediction on Soft Clay Foundation

Machine learning-aided selection of CPT-based transformation models using field monitoring data from a specific project