Decomposition approach to solve dial-a-ride problems using ant Computing and Constraint Programming

Abstract

In this paper we solve the Dial-A-Ride Problem (DARP). The main objective of the DARP is to minimize operation costs for renting pieces of work from the transportation service providers. The resolution approach considered in this work, starting from a network formulation of the DARP, decomposes the problem in two phases: Clustering and Chaining. We model both phases like a Set Partitioning Problem (SPP) and solve them with an interesting sinergy between two different optimization methods: Ant Computing and Constraint Programming. © 2007 Springer-Verlag Berlin Heidelberg.