Bi and tri-objective optimization in the deterministic network interdiction problem

Abstract

Solution approaches to the deterministic network interdiction problem have previously been developed for optimizing a single figure-of-merit of the network configuration (i.e. flow that can be transmitted between a source node and a sink node for a fixed network design) under constraints related to limited amount of resources available to interdict network links. These approaches work under the assumption that: (1) nominal capacity of each link is completely reduced when interdicted and (2) there is a single criterion to optimize. This paper presents a newly developed evolutionary algorithm that for the first time allows solving multi-objective optimization models for the design of network interdiction strategies that take into account a variety of figures-of-merit. The algorithm provides an approximation to the optimal Pareto frontier using: (a) techniques in Monte Carlo simulation to generate potential network interdiction strategies, (b) graph theory to analyze strategies' maximum source-sink flow and (c) an evolutionary search that is driven by the probability that a link will belong to the optimal Pareto set. Examples for different sizes of networks and network behavior are used throughout the paper to illustrate and validate the approach.