Hybrid evolutionary algorithm with Hermite radial basis function interpolants for computationally expensive adjoint solvers
Hybrid evolutionary algorithm with Hermite radial basis function interpolants for computationally expensive adjoint solvers
In this paper, we present an evolutionary algorithm hybridized with a gradient-based optimization technique in the spirit of Lamarckian learning for efficient design optimization. In order to expedite gradient search, we employ local surrogate models that approximate the outputs of a computationally expensive Euler solver. Our focus is on the case when an adjoint Euler solver is available for efficiently computing the sensitivities of the outputs with respect to the design variables. We propose the idea of using Hermite interpoloation to construct gradient-enhanced radial basis function networks that incorporate sensitivity data provided by the adjoint Euler solver. Further, we conduct local search using a trust-region framework that interleaves gradient-enhanced surrogate models with the computationally expensive adjoint Euler solver. This ensures that the present hybrid evolutionary algorithm inherits the convergence prperties of the classical trust-region approach. We present numerical results for airfoil aerodynamic design optimization problemss to show that the proposed algorithm converges to good designs on a limited computational budget.
hybrid evolutionary algorithm, hermite radial basis function, gradient-based approximation, computationally expensive adjoint solver
97-119
Ong, Y.S.
62497a6f-823e-4663-b263-4a805a00f181
Lum, K.Y.
d37541a4-936a-49bf-a4a2-711fb2b54512
Nair, P.B.
d4d61705-bc97-478e-9e11-bcef6683afe7
13 September 2008
Ong, Y.S.
62497a6f-823e-4663-b263-4a805a00f181
Lum, K.Y.
d37541a4-936a-49bf-a4a2-711fb2b54512
Nair, P.B.
d4d61705-bc97-478e-9e11-bcef6683afe7
Ong, Y.S., Lum, K.Y. and Nair, P.B.
(2008)
Hybrid evolutionary algorithm with Hermite radial basis function interpolants for computationally expensive adjoint solvers.
Computational Optimization and Applications, 39 (1), .
(doi:10.1007/s10589-007-9065-5).
Abstract
In this paper, we present an evolutionary algorithm hybridized with a gradient-based optimization technique in the spirit of Lamarckian learning for efficient design optimization. In order to expedite gradient search, we employ local surrogate models that approximate the outputs of a computationally expensive Euler solver. Our focus is on the case when an adjoint Euler solver is available for efficiently computing the sensitivities of the outputs with respect to the design variables. We propose the idea of using Hermite interpoloation to construct gradient-enhanced radial basis function networks that incorporate sensitivity data provided by the adjoint Euler solver. Further, we conduct local search using a trust-region framework that interleaves gradient-enhanced surrogate models with the computationally expensive adjoint Euler solver. This ensures that the present hybrid evolutionary algorithm inherits the convergence prperties of the classical trust-region approach. We present numerical results for airfoil aerodynamic design optimization problemss to show that the proposed algorithm converges to good designs on a limited computational budget.
Text
Ong_08.pdf
- Accepted Manuscript
More information
Submitted date: 14 February 2005
Published date: 13 September 2008
Keywords:
hybrid evolutionary algorithm, hermite radial basis function, gradient-based approximation, computationally expensive adjoint solver
Identifiers
Local EPrints ID: 64451
URI: http://eprints.soton.ac.uk/id/eprint/64451
ISSN: 0926-6003
PURE UUID: deb8eaf4-c550-48a6-95bf-14d4f5715480
Catalogue record
Date deposited: 24 Dec 2008
Last modified: 15 Mar 2024 11:49
Export record
Altmetrics
Contributors
Author:
Y.S. Ong
Author:
K.Y. Lum
Author:
P.B. Nair
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics