Quantifying the evolution of settlement risk for surrounding environments in underground construction via complex network analysis

• Published in: Tunnelling and underground space technology Vol. 103; p. 103490
• Main Authors: Zhou, Cheng, Kong, Ting, Jiang, Shuangnan, Chen, Shida, Zhou, Ying, Ding, Lieyun
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.09.2020; Elsevier BV
• Subjects: Complex network; Construction sites; Evolution; Network analysis; Occupational safety; Risk; Risk assessment; Risk factors; Safety; Spatial distribution; Underground construction; Underground surrounding environment; Windows (intervals); Wuhan China; Risk assessment; Underground surrounding environment; Complex network
• ISSN: 0886-7798; 1878-4364
• DOI: 10.1016/j.tust.2020.103490

•Complex network is used to evaluate settlement risk in underground construction.•Temporal evolution of risk is measured by network entropy with time windows.•Spatial distribution of settlement risk is qualified by network structure.•The risk assessment for surrounding environments is verified in a real metro project. The rapid development of underground construction results in safety problems in surrounding environments due to ground surface settlement. Comparing accumulated settlement with threshold to evaluate settlement risk is insufficiently accurate. Thus, identifying the location and change in the settlement risk is difficult. This research develops a data-based complex network approach to analyze and assess the risks of surrounding environments from a synthetic and dynamic perspective based on settlement data and address these limitations. In the proposed approach, a complex network is divided into multiple time windows, from which temporal dynamics of settlement can be captured. Then, betweenness centrality is used to describe the spatial distributions of settlement risk, and network entropy is used to quantify the evolution of settlement risk. A case of Wuhan metro project in China is used to validate the effectiveness and feasibility of the developed approach. Results demonstrated that the developed approach can reveal spatio–temporal evolution law of settlement risks of surrounding environments. The developed method not only offers a systematic way to interpret settlement risk of surrounding environment in underground construction but also has the potential to improve safety performance in construction sites.