Enhanced safety prediction of vault settlement in urban tunnels using the pair-copula and Bayesian network

To analyze and study the key control factors of the risk of vault deformation in tunnels excavated at small clear distances, a risk assessment method based on a pair-copula Bayesian network (PCBN) model is proposed. Based on an analysis of the factors affecting the settlement of the vault of an excavated tunnel, an index system for the risk assessment of the vault settlement is established. A PCBN model based on a pair-copula function and BN can effectively deal with the complex risk system and the correlation problems within the risk system. Using the constructed PCBN model, a risk analysis of the dome settlement caused by the excavation in the Donghu Tuanshan tunnel in Wuhan is carried out, and the risk status of the dome settlement of the tunnel is determined to be basically safe. It can be determined from this research that a pair-copula correlation concept combined with a BN retains the strength of both concepts, and a safety risk analysis method for the settlement of tunnel vaults based on the PCBN model is proposed to perform real-time and effective safety risk assessment and provide decision support for the shield construction stage. Through a correlation analysis of the risk indicators of the risk system, it is determined that the risk indicators that have the most significant impact on the deformation of the vault are the state of the groundwater, the complexity of the construction environment, the soil quality of the tunnel arch bottom, and the soil quality of the tunnel vault. These risk indicators are used as key risk indicators for decision-making. Based on this method, decision -making suggestions for reducing the risk of vault deformation are proposed. Good results are achieved in project implementation, and safety management of the vault deformation of a small clear excavated tunnel is provided. & COPY; 2022 Elsevier B.V. All rights reserved.

Automated summary

A risk assessment method for vault deformation in small clear distance tunnels is proposed using a pair-copula Bayesian network (PCBN) model. An index system for risk assessment is established based on factors affecting vault settlement. The PCBN model effectively addresses the complexity and correlation within the risk system. The model is applied to analyze the risk status of dome settlement in the Donghu Tuanshan tunnel, and it is determined to be safe. The method provides real-time and effective safety risk assessment and decision support for shield construction.