Efficient accelerators for PSO in an inverse design of multi-element airfoils

Abstract

The main object in aerodynamic design is a wing. High lift (HL) systems are probably the main concern in wing design optimization due to strict regulations. An optimization of high lift systems is difficult because of flow physics and high number of design parameters. Evolutionary algorithms including particle swarm optimization (PSO) are popular methods in HL system optimization. However, the computational burden of PSO may be a serious issue in process. In this article, efficient accelerators are integrated to PSO and applied to an inverse design of multi-element airfoils. The comparative test cases demonstrated that remarkable reductions in computational times have been accomplished. (C) 2019 Elsevier Masson SAS. All rights reserved.