Skip to Main content Skip to Navigation

Étude de l'hybridation des méta-heuristiques, application à un problème d'ordonnancement de type jobshop

Abstract : In this thesis, iterative search methods are studied and applied to the jobshop scheduling problem. Instead of trying to reach high performance, we focus our work towards a better understanding of search algorithms and their hybridization. In particular, we study the impact of hybridization on the quality of final solutions.

First part of our work consists in evaluating the impact of combining several criteria in the objective function. We have used several local search algorithms in order to gather statistical informations about the resulting structures of fitness landscape. We have also studied the impact on performance.

Then, the study of the interactions between operators in our GA leads us to define some improvements of the scheme of application of genetic-operators.

Finally, we study several schemes of hybridization based on tabu-search and evolutionary algorithms.
Document type :
Theses
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00008729
Contributor : David Duvivier <>
Submitted on : Tuesday, March 8, 2005 - 7:27:02 PM
Last modification on : Friday, July 19, 2019 - 11:28:01 AM
Document(s) archivé(s) le : Friday, April 2, 2010 - 9:15:39 PM

Identifiers

  • HAL Id : tel-00008729, version 1

Collections

Citation

David Duvivier. Étude de l'hybridation des méta-heuristiques, application à un problème d'ordonnancement de type jobshop. Autre [cs.OH]. Université du Littoral Côte d'Opale, 2000. Français. ⟨tel-00008729⟩

Share

Metrics

Record views

711

Files downloads

638