The undirected m-Capacitated Peripatetic Salesman Problem

Éric Duchenne (1) , Gilbert Laporte (2) , Frédéric Semet (3)

1 LAMIH - Laboratoire d'Automatique, de Mécanique et d'Informatique industrielles et Humaines - UMR 8201
2 CIRRELT - Laboratoire CIRRELT Université Laval Quebec
3 LAGIS-OSL

In the m-Capacitated Peripatetic Salesman Problem (m-CPSP) the aim is to determine m Hamiltonian cycles of minimal total cost on a graph, such that all the edges are traversed less than the value of their capacity. This article introduces three formulations for the m-CPSP. Two branch-and-cut algorithms and one branch-and-price algorithm are developed. Tests performed on randomly generated and on TSPLIB Euclidean instances indicate that the branch-and-price algorithm can solve instances with more than twice the size of what is achievable with the branch-and-cut algorithms.

Mots clés: Peripatetic Salesman Problem; Traveling Salesman Problem; Capacity; Branch-and-cut; Branch-and-price

Domaines: Recherche opérationnelle [math.OC]

Éric Duchenne, Gilbert Laporte, Frédéric Semet. The undirected m-Capacitated Peripatetic Salesman Problem. European Journal of Operational Research, 2012, 223 (3), pp.637-643. ⟨10.1016/j.ejor.2012.07.022⟩. ⟨hal-00831917⟩