A Hierarchical Genetic Algorithm Using Multiple Models for Optimization

摘要: This article presents both the theoretical basis and some experimental results on Hierarchical Genetic Algorithms (HGAs). HGAs are explained in details, along with advantages conferred by their multi-layered hierarchical topology. topology is an excellent compromise to classical exploration/exploitation dilemma. Another feature introduction of multiple models for optimization problems, within frame HGA. We show that such architecture, it possible use a mix simple very fast more complex (with slower solvers), still achieve same quality as obtained only models. The different concepts presented this paper then illustrated via experiments Computational Fluid Dynamics problem, namely nozzle reconstruction. overall Algorithm using can classic GA model, but up three times faster.