Vehicle Routing Problem with Uncertain Costs via a Multiple Ant Colony System

We consider the capacitated vehicle routing problem (VRP) withuncertain travel costs, where the uncertainty represents the realistic factors like unfriendly weather conditions, traffic jams, etc. In this chapter, we present a multiple ant colony system approach in which ant colony optimization processes work concurrently to produce multiple solutions. Experimental results on different types of VRP instances (with clustered customers, randomly placed customers, and a mixed of the previous two), each instance having 200 customers, are finally discussed