Assessing road network vulnerability: A user equilibrium interdiction model
Stefano Starita, Maria Paola Scaparra
Abstract
Road networks are vulnerable to natural and man-made disruptions. The loss of one or many critical links of the network often leads to increased traffic congestion. Therefore, quantitative models are necessary to identify these critical assets so that actions can be taken by decision makers to mitigate the impact of disruptions. This paper proposes an optimisation model to identify the set of arcs that, when lost, results in the worst congestion under user equilibrium traffic. The model is formulated as a bi-level non-linear problem. The challenging formulation is solved via a customised version of Greedy Randomised Adaptive Search Procedure (GRASP) meta-heuristic. Computational experiments are run on a dataset of artificial grids and managerial insights are provided based on popular Sioux and Berlin network case-studies.