Gradient-Enhanced Metamodels and Multiparametric Strategies for Designing Structural Assemblies
Résumé
This paper proposes a multilevel model optimization strategy for structural assemblies. The general objective is to reduce computation costs; here, we focus on the costs which are associated with the generation of metamodels. Our goal is achieved through the introduction of two main elements: the multiparametric Strategy based on the LATIN method, which reduces the computation costs when the parameters vary, and the use of gradient-based metamodels. Cokriging and radial basis functions (RBF) metamodels are presented and performance of these approximations is illustrated with analytical and mechanical examples with one to four design variables.