Titre :

Two-Level Algorithm Combining Bayesian Optimization and Swarm Intelligence for Variable-Size Optimal Layout Problems

Auteur(s) :

Gamot, Juliette [Auteur]
Centre Inria de l'Université de Lille
DTIS, ONERA, Université Paris Saclay [Palaiseau]
Balesdent, Mathieu [Auteur]
DTIS, ONERA, Université Paris Saclay [Palaiseau]
Tremolet, Arnault [Auteur]
DTIS, ONERA, Université Paris Saclay [Palaiseau]
Wuilbercq, Romain [Auteur]
DTIS, ONERA, Université Paris Saclay [Palaiseau]
Melab, Nouredine [Auteur]
Optimisation de grande taille et calcul large échelle [BONUS]
Talbi, El-Ghazali [Auteur]
Centre Inria de l'Université de Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Université de Lille
Optimisation de grande taille et calcul large échelle [BONUS]

Titre de la manifestation scientifique :

GECCO '23 Companion: Companion Conference on Genetic and Evolutionary Computation

Ville :

Lisbonne

Pays :

Portugal

Date de début de la manifestation scientifique :

2023-07-15

Éditeur :

ACM

Date de publication :

2023-07-24

Mot(s)-clé(s) :

optimisation bayésienne
intelligence en essaim
optimisation multi-échelle

Mot(s)-clé(s) en anglais :

bayesian optimization
swarm intelligence
multilevel optimization

Discipline(s) HAL :

Informatique [cs]
Sciences de l'ingénieur [physics]

Résumé en anglais : [en]

The design process of complex engineering systems may involve problems in which the number and type of design variables and constraints vary throughout the optimization process based on the values of dimensional variables. ...
Lire la suite >The design process of complex engineering systems may involve problems in which the number and type of design variables and constraints vary throughout the optimization process based on the values of dimensional variables. This category of problems is called Variable-Size Design Space optimization. A well-known application is the optimal layout problem which requires to place a variable number of components into a container. The dual objective is here to optimize the list of components in addition to their placements within the container. In this paper, a two-level algorithm is described to solve the aforementioned optimal layout problems. This algorithm combines the strength of a Swarm Intelligence algorithm based on a virtual-force system and a discrete Bayesian Optimization algorithm purposely adapted to tackle the dimensional aspect of this problem. The implementation of the two-level algorithm is discussed and the proposed approach is applied to the layout optimization of a satellite module. The performance of the algorithm is then analyzed with respect to several occupation rates of the system. This approach is also compared with a Hidden-Genes Genetic Algorithm.Lire moins >

Langue :

Anglais

Comité de lecture :

Oui

Audience :

Internationale

Vulgarisation :

Non

Collections :

• Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL