Study of Research Progress on Prediction Methods for Safety Performance of Shield Tunnels

With the rapid development of urban underground space utilization, shield tunnel engineering is confronting increasingly complex geological conditions and environmental effects. The prediction of safety performance has become a critical technology for ensuring the entire life cycle safety of projects. In recent years, various prediction methods have developed through mutual integration, but there is still a lack of systematic classification and comparative analysis. This article comprehensively reviews the main method systems for predicting the safety performance of shield tunnels, including numerical simulation, model testing, mechanical analytical models, machine learning, and grey theory, etc. It systematically analyzes the theoretical basis, application characteristics, and limitations of each method, clearly explains the applicability of each method, and looks forward to the future development trend of multi-method integration. The aim is to provide a systematic reference for related research.