Exploring spatio-temporal correlation and complexity of safety monitoring data by complex networks

• Published in: Automation in construction 2022-03, Vol.135, p.104115, Article 104115
• Main Authors: Gao, Yuyue, Li, Rao, Zhou, Cheng, Jiang, Shuangnan
• Format: Article
• Language: English
• Subjects: Complex network; Complexity; Data collection; Eigenvectors; Metro construction; Monitoring; Nodes; Risk assessment; Safety; Settlement time series; Spatio-temporal correlation; Subway construction; Topological analysis; Wavelet transform
• ISSN: 0926-5805; 1872-7891
• DOI: 10.1016/j.autcon.2021.104115

With the development of monitoring technologies, collected data has become more massive, precise, and timely, which can be an excellent foundation for more comprehensive assessment. However, existing data-analyzing researches still concentrate mainly on spatial or temporal characteristics separately, neglecting the dependence inside. Actually, the spatio-temporal correlation has been widely studied in other areas. Based on that, this paper proposed an undirected and unweighted data-based complex network as a risk assessment tool to explore the spatio-temporal correlation in safety monitoring data. Eigenvectors containing both spatial and temporal characteristic values are the nodes and the degree of correlation determines whether edges exist between each pair of nodes. The good application of the model in a metro construction project verifies the existence and significance of the correlation. This work not only reveals the spatio-temporal correlation of construction characteristics but also provides a new perspective in safety assessment. •Spatio-temporal feature of safety monitoring data is extracted.•Complex networks are constructed based on spatio-temporal correlation.•The complexity of safety monitoring data is revealed with the topological structure of complex network.•An actual case of Wuhan metro construction is introduced to discuss the proposed theory under practical circumstances.