Connexion utilisateur 0 Panier
Accueil Livres eBooks Revues spécialisées Références et comptes rendus Auteurs, éditeurs, examinateurs Index A-Z des produits
Flexible Automation and Intelligent Manufacturing, 1997:
Proceedings of the Seventh International FAIM Conference

ISBN Imprimer: 978-1-56700-089-4

ISBN En ligne: 978-1-56700-442-7

PLUG AND PLAY FRAMEWORK FOR COMBINATORAL PROBLEM HEURISTICS

Résumé

Combinatorial problems such as the traveling salesperson, assignment and set-covering problems often rear their heads in process optimization. Because these problems are NP-hard, it remains unlikely that they will yield global optima at reasonable computational cost. Therefore, we must rely on heuristic techniques which seek good, albeit sub-optimal, solutions at an acceptable cost. To this end, meta-heuristic search strategies are interesting because they require very little problem specific knowledge. What little they do need to know about a problem can be encapsulated in a neighborhood operator which provides a mechanism for seeking local optima. Unfortunately, these operators can be very difficult to implement and in this paper we define a practical interface for enumerating neighbors. Once a neighborhood operator has been defined, dynamic strategy selection becomes possible. This means that meta-heuristic strategies, such as simulated annealing and tabu search, can be plugged into problem contexts at run-time.
Accueil Portail numérique Begell Bibliothèque numérique Begell Revues spécialisées Livres eBooks Références et comptes rendus Auteurs, éditeurs, examinateurs Index A-Z des produits Prix et politiques d'abonnement A propos de Begell House Contactez-nous Language English 中文 Русский 日本語 Português Deutsch Français Español