A Bi-objective metaheuristic for disaster relief operation planning

Abstract

In this chapter we consider a multi-objective problem arising in a post-natural-disaster situation. It is assumed that the infrastructure in the affected region has been partly destroyed by an earthquake, flood or tsunami. The problem treated in the following concerns survival help which means the supply of aliment, shelter and medicine to the population living in the affected area, compensating for the destroyed facilities. Therefore a multi-objective covering tour problem has to be faced, that aims at distributing the mentioned items among the population. We developed a hybrid method based on genetic algorithms, variable neighborhood search and path relinking to solve the real-world motivated problem. The algorithm is tested on real world data from the province Manabí in Ecuador. The results on small instances are compared to an epsilon constraint method. Also results for medium- and large-size real-world instances are presented. © 2010 Springer-Verlag Berlin Heidelberg.