Technical gazette, Vol. 32 No. 6, 2025.
Original scientific paper
https://doi.org/10.17559/TV-20241101002107
A Novel Subway Tunnel Collapse Risk Prediction Model using Mutation Theory
Tao Ma
; School of Management, School of Civil Engineering, Shijiazhuang TieDao University, Shijiazhuang 0500543, HeBei, China
Xiaodong Li
; CCCC Third Highway Engineering Co., Ltd, Beijing, Shunyi Distric,101300
Tianjuan Gao
; Department of Logistics Management, Hebei Jiaotong Vocational and Technical College, Shijiazhuang 050035, HeBei, China
*
* Corresponding author.
Abstract
With the rapid expansion of urbanization, subway tunnel construction has grown significantly in scale and speed, but frequent collapse incidents pose severe risks to safety, economy, and public infrastructure. This study proposes a novel risk prediction model for subway tunnel collapse using mutation theory, which effectively captures the dynamic and discontinuous nature of collapse events. The model integrates geological, hydrological, and construction parameters and employs real-time monitoring data to achieve dynamic risk classification. Numerical simulations indicate a strong correlation between settlement, clearance convergence, and deformation, with the top settlement reaching 0.28 mm and plastic strain increasing from 1.3 × 10−4 to 1.23 × 10−3 after excavation. Case verification reveals high collapse risks at multiple monitoring points, underscoring the importance of robust monitoring and preventive strategies. The findings demonstrate that mutation theory-based risk modeling is feasible and reliable, providing a new theoretical framework and practical tool for early warning and risk management in subway tunnel construction. Future research will explore the model's applicability to diverse geological conditions and its integration with advanced monitoring systems to enhance prediction accuracy and risk mitigation.
Keywords
collapse risk estimation; geological features; hydrologic condition; mutation theory; subway; tunnel
Hrčak ID:
337718
URI
Publication date:
31.10.2025.
Visits: 110 *