Flexible variable neighborhood search in dynamic vehicle routing

Abstract

Many optimization problems are dynamic, which means that the available data may change while the problem is being solved. Incorporating elements into the algorithm that take into account these changes usually leads to more effective algorithms which provide better solutions. In this work, we propose a flexibility strategy for the Vehicle Routing Problem with Dynamic Requests. We show that early decissions, which are taken in the beginning of the optimization process, influence the quality of final solutions for the dynamic problem. Our flexible algorithm provides better results than the canonical one and is competitive with the results in the literature. © 2011 Springer-Verlag.