A simple and efficient pre-selection method for partial network scan in critical link identification

Transportation networks are highly susceptible to disruptions, with certain segments significantly impacting overall network functionality. Identifying these vulnerable parts, known as critical links, is essential for vulnerability analysis. Numerous studies have addressed vulnerability analysis, differing primarily in the performance measures employed and the identification approach adopted. While a full network scan is considered the most accurate, it becomes computationally intensive for larger networks, especially when operational aspects are considered. Although topological measures are feasible, they do not fully capture the network’s performance characteristics. This paper presents simple and easily implementable pre-selection criteria to identify candidate links. Though preselection-based methods are already available in literature, these are often limited by cumbersome implementation and inaccuracy. The criteria proposed in this paper leverage betweenness centrality, relative size of giant component, and link positions to select candidate links for vulnerability analysis. The network scan is applied exclusively to these pre-selected links, reducing computation time while maintaining accuracy. Experiments were conducted on eight different networks to validate the approach. The results reinforce the credibility of the pre-selection criteria, demonstrating their efficacy in identifying critical links. The methodology also performed effectively even when tested under different demand and supply levels within the same network. In conclusion, the proposed pre-selection criteria offer a practical and time-efficient solution for identifying critical links in transportation networks.