Machine learning-based forecasting of ground surface settlement induced by metro shield tunneling construction

Area of Science:

• Geotechnical Engineering
• Civil Engineering
• Urban Planning

Background:

• Urban rail transit systems, like subways, are vital for managing urbanization and traffic congestion.
• Shield tunneling, a common subway construction method, causes ground surface settlement, posing risks to adjacent buildings.
• Accurate prediction of ground settlement is essential for safety warnings and evaluations during shield tunneling.

Purpose of the Study:

• To develop a reliable method for predicting ground surface settlement induced by shield tunneling.
• To enhance the accuracy of settlement prediction using a combination of data preprocessing and optimized neural networks.
• To provide a valuable reference for safety management and control of ground settlement.

Main Methods:

• Collected multi-point ground surface settlement data from monitoring sections.
• Applied a tangent circle-based data preprocessing method to convert discrete data into continuous, smooth data.
• Utilized the Particle Swarm Optimization (PSO) algorithm to optimize a Back Propagation Neural Network (BPNN) for settlement prediction.
• Investigated the impact of network structure and prediction modes on prediction accuracy.

Main Results:

• The proposed data preprocessing and PSO-BP algorithm achieved a maximum relative error of 0.46% in settlement prediction.
• The method demonstrated a good fitting effect, accurately predicting settlement over a 5-day period during both slow and stable settlement stages.
• The study validated the predictive performance and effectiveness of the integrated approach.

Conclusions:

• The developed method provides accurate and reliable predictions of ground surface settlement caused by shield tunneling.
• This approach offers a valuable tool for real-time safety warnings and effective control of ground settlement during subway construction.
• The findings contribute to enhancing the safety and efficiency of urban underground infrastructure development.