Surrogate-based agents for constrained optimization

Diane Villanueva (1) , Rodolphe Le Riche (2, 3, 4, 5, 6) , Gauthier Picard (7, 4, 5, 8) , Raphael T. Haftka (1)

1 UF|MAE - Department of Mechanical and Aerospace Engineering [Gainesville]
2 CROCUS-ENSMSE - Equipe : Calcul de Risque, Optimisation et Calage par Utilisation de Simulateurs
3 DEMO-ENSMSE - Département Décision en Entreprise : Modélisation, Optimisation
4 Mines Saint-Étienne MSE - École des Mines de Saint-Étienne
5 FAYOL-ENSMSE - Institut Henri Fayol
6 LIMOS - Laboratoire d'Informatique, de Modélisation et d'Optimisation des Systèmes
7 ISCOD-ENSMSE - Département Informatique pour les Systèmes Coopératifs Ouverts et Décentralisés
8 LHC - Laboratoire Hubert Curien

Multi-agent systems have been used to solve complex problems by decomposing them into autonomous subtasks. Drawing inspiration from both multi-surrogate and multi-agent techniques, we de ne in this article optimization subtasks that employ di erent approxi- mations of the data in subregions through the choice of surrogate, which creates surrogate- based agents. We explore a method of design space partitioning that assigns agents to subregions of the design space, which drives the agents to locate optima through a mixture of optimization and exploration in the subregions. These methods are illustrated on two constrained optimization problems, one with uncertainty and another with small, discon- nected feasible regions. It is observed that using a system of surrogate-based optimization agents is more e ective at locating the optimum compared to optimization with a single surrogate over the entire design space.

Domaines: Système multi-agents [cs.MA]; Optimisation et contrôle [math.OC]

Diane Villanueva, Rodolphe Le Riche, Gauthier Picard, Raphael T. Haftka. Surrogate-based agents for constrained optimization. 14th AIAA Non-Deterministic Approaches Conference, Apr 2012, Honolulu, United States. ⟨10.2514/6.2012-1935⟩. ⟨emse-00680732⟩