Link criticality identification and analysis in road networks using path flow weighted betweenness centrality

The hybrid measures integrate network topology and traffic attributes to evaluate link criticality, balancing assessment accuracy and computational efficiency. Existing studies mainly use link traffic attributes to weight topology-based measures forming hybrid measures applied to identify critical single links. Link flow weighted betweenness centrality is a typical hybrid measure. However, this measure may lead to these issues: (i) zero-flow paths can contribute to link criticality; (ii) low-flow paths have the same contribution to link criticality with high-flow paths; (iii) inability to identify critical link combinations. In this study, we refine BC in road networks and take path flow as its weight, addressing issues (i) and (ii). Besides, we extend the refined measures from identifying critical single links to identifying critical link combinations and propose an iterative optimization method to identify the top K critical link combinations. The measures and methods are tested on three benchmark networks, demonstrating their superiority and validity in ranking consistency and computational efficiency. Our findings provide valuable support for traffic managers in identifying and analyzing critical links in large-scale road networks, helping them develop effective strategies to reduce disaster risks and enhance network resilience.