Multi-source information fusion for safety risk assessment in underground tunnels

• Main Authors: Guo, Kai, Zhang, Limao
• Other Authors: School of Civil and Environmental Engineering
• Format: Article
• Language: English
• Published: 2022
• Subjects: Engineering::Civil engineering; Information Fusion; Risk Assessment
• Online Access: https://hdl.handle.net/10356/160690
• Institution: Nanyang Technological University

Risk management has become one of the most important issues in the underground tunnel construction due to the rapidly increasing scale. A hybrid approach integrating Building Information Modeling (BIM) and the Dempster Shafer (D–S) evidence theory is proposed to support systematic risk assessment and visualization in underground tunnels. BIM is used to build three dimensional (3D) models, an application programming interface (API) to extract the engineering information, the D–S evidence theory to fuse information and determine the risk probability, Dynamo to realize real-time visualization, and an evidence updating method to capture the dynamic features of the risk status. A cross-river tunnel case in the city of Wuhan, China, is used to test the effectiveness and applicability of the proposed approach. It is found that (1) Three target tunnel sections are determined as under safe, low risk, and low risk levels, respectively; (2) The defect of design variables is the main factor leading the tunnel sections to unsafe levels; (3) Dynamics of the tunnel condition can be captured by the incorporation of the evidence updating method, in which higher certainty and reliability are demonstrated. The novelty of the proposed approach lies in (a) combining the advantages of BIM for dynamic data processing with the capabilities of the D–S evidence theory for information fusion; (b) an evidence updating method is incorporated to capture the dynamic of the tunnel construction. This hybrid approach is expected to enrich the risk management for complex underground projects by fusing multi-source information subjected to uncertainty, conflicts, and dynamics