A multi-depot heterogeneous vehicle routing problem in cities affected by a large scale disaster - Université de technologie de Troyes Accéder directement au contenu
Communication Dans Un Congrès Année : 2015

A multi-depot heterogeneous vehicle routing problem in cities affected by a large scale disaster

Résumé

This study takes place in a national project on disaster logistics and concerns the last mile distribution phase after a large scale disaster, taking the Haiti earthquake as case study. From supply depots in the suburbs, the goal is to design vehicle routes with multiples trips to reach camps of refugees, knowing that the streets can be more or less obstructed, which implies several types of vehicles. The goal is to minimize the total distribution time and the cost of vehicles used. The proposed solution method is a matheuristic derived from Penna et al. (2013), a multi-start iterated local search. The improvement procedure is a variable neighborhood search (VND). A set-partitioning problem (SPP) is solved periodically, using a pool of good routes found by the heuristic, but the communication is bidirectional: the heuristic is also called each time a new solution is obtained during the SPP resolution. The algorithm competes with published metaheuristics on multi-depot VRP instances from the literature. On data from Haiti earthquake (16,000 nodes, 19,000 edges, 12 depots, 62 camps, 3 vehicle types), the algorithm returns in one minute on a PC solutions which are considered as very good by the decision maker.
Fichier non déposé

Dates et versions

hal-02292153 , version 1 (19-09-2019)

Identifiants

  • HAL Id : hal-02292153 , version 1

Citer

Vaz Penna Puca Huachi, Andréa Cynthia Santos, Christian Prins. A multi-depot heterogeneous vehicle routing problem in cities affected by a large scale disaster. 5th Workshop on Combinatorial Optimization, Routing and Location (CORAL 2015), Sep 2015, Salamanca, Spain. ⟨hal-02292153⟩

Collections

CNRS UTT LOSI
31 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More