multiobjective optimization in hydrodynamic stability control

Frank Strauss (1) , Jean-Antoine Desideri (2) , Régis Duvigneau (2) , Vincent Heuveline (1)

1 KIT - Karlsruhe Institute of Technology
2 OPALE - Optimization and control, numerical algorithms and integration of complex multidiscipline systems governed by PDE

We focus on a shape optimization problem in hydrodynamic stability control. The goal is to minimize the drag in a flow regime by modification of the underlying geometry while ensuring hydrodynamic stability. We propose an extension of a numerical scheme to speed up the computations and avoid the solving of the original PDE model. Therefore, a metamodel is used as approximation of the original model. A Particle Swarm Optimization algorithm is used in conjunction with a Nash game to consider both objectives simultaneously, with a split of design variables established from a sensitivity analysis accounting for first and second derivatives of functionals related to the metamodel.

Mots clés: Optimum shape design; hydrodynamic stability; metamodel; particle swarm optimization; Nash game

Domaines: Analyse numérique [cs.NA]

Frank Strauss, Jean-Antoine Desideri, Régis Duvigneau, Vincent Heuveline. multiobjective optimization in hydrodynamic stability control. [Research Report] RR-6608, INRIA. 2008. ⟨inria-00309693⟩