The stochastic approach in road network vulnerability analysis

This paper draws attention to the vulnerability of road networks, when larger numbers of links are interrupted. These situations can arise as a consequence of natural disasters, terrorist attacks or additional random events. The primary aim of the vulnerability analysis is to identify those links whose interruptions have the highest impact on the given network. In order to identify the respective links, a large number of combinations of the links need to be analyzed which is beyond the scope of any computer. In this paper we provide a means of dealing with the problem using methods based on simulated annealing which do not produce the worst-case scenario in general but which enable us to produce scenarios with a sufficiently high impact.