Detection of critical links in spatial-temporal routes based on complex networks
DOI 10.51955/2312-1327_2025_3_Β
Abstract: This paper presents an enhanced method for detecting critical edges in spatial-temporal route networks based on complex network analysis. Unlike previous models, the proposed approach accounts not only for the topological characteristics of routes but also for their dynamic variability through a composite weight that includes the average travel speed and its variance. Additional metrics, such as load centrality and robustness of connectivity, are introduced. Critical edges are automatically identified via percolation analysis, without the need for manual threshold adjustment. For criticality prediction, a gradient boosting model is employed, using a set of structural and temporal features. The proposed approach ensures more accurate, reproducible, and adaptive identification of vulnerable segments in route networks and can be applied in real time to support air traffic management.
Key words: air transportation, airway network, complex networks theory, spatial-temporal network, critical link detection, network centrality, percolation theory.
Download article in PDF 1.6 MB