Publication detail
Surrogate-based Shape Optimization in Aerodynamic Applications
DVOŘÁK, P.
Czech title
Aerodynamická tvarová optimalizace s využitím metamodelů
English title
Surrogate-based Shape Optimization in Aerodynamic Applications
Type
journal article - other
Language
en
Original abstract
A preliminary aerodynamic design often imposes requirements on global optimum search within a large, highly multimodal design space. Tools typically deployed to evaluate individual design candidates are very computationally expensive, being part of the finite volume computational fluid dynamics class. This virtually prevents deployment of traditional stochastic global optimization approaches, such as evolutionary algorithms. Hence, there has been a growing interest in metamodelling techniques, providing a reliable surrogate of the simulator response to an optimization algorithm. Efficient deployment of such techniques within preliminary aerodynamic design is of interest to Garteur Action Group 52 members. The present paper describes the involvement of Brno University of Technology, Institute of Aerospace Engineering in the AG52. The considered test case is based on the RAE2822 aerofoil constrained multipoint optimization problem. The overall problem setup is given along with selected surrogate modelling and optimization techniques. The presented approach featuring artificial neural networks is able to produce highly reliable metamodels with cutting-edge performance as documented by the AG52 performance metrics comparison. The work described within this paper has been presented at EUROGEN 2015 conference in Glasgow.
Czech abstract
Koncepční aerodynamický návrh klade nároky na nalezení globálního optima ve velkém, vysoce multimodálním návrhovém prostoru. Nástroje užívané k evaluaci jednotlivých návrhových vektorů jsou typicky značně výpočetně náročné. To prakticky znemožnuje aplikaci tradičních stochastických globálních optimalizačních metod jako jsou např. evoluřní algoritmy. Z výše uvedeného důvodu se zvedá zájem o optimalizační algoritmy založené na využití náhradních modelů – metamodelů. Metamodely poskytují kvalitní aproximace skutečné odezvy modelu pro optimalizační algoritmus. Účinným využitím metamodelů pro koncepční aerodynamickou optimalizaci se zabývají členové pracovní skupiny 52 mezinárodní organizace Garteur. Předkládaný text popisuje výsledky Leteckého ústavu z VUT v Brně dosažené v rámci zapojení do AG52. Vyvinutá metodika je demonstrována na optimalizaci profilu RAE2822 se čtrnácti stupni volnosti. Jedná se o případ multikriteriální optimalizace s geometrickými i aerodynamickými omezeními. Vyvinutá metodika využívá umělých neuronových sítí a generuje spolehlivé metamodely, které dosahují ve srovnání s ostatními přístupy v AG52 špičkových výsledků. Výsledky popisované v tomto textu byly předneseny na konferenci EUROGEN 2015 v Glasgow.
English abstract
A preliminary aerodynamic design often imposes requirements on global optimum search within a large, highly multimodal design space. Tools typically deployed to evaluate individual design candidates are very computationally expensive, being part of the finite volume computational fluid dynamics class. This virtually prevents deployment of traditional stochastic global optimization approaches, such as evolutionary algorithms. Hence, there has been a growing interest in metamodelling techniques, providing a reliable surrogate of the simulator response to an optimization algorithm. Efficient deployment of such techniques within preliminary aerodynamic design is of interest to Garteur Action Group 52 members. The present paper describes the involvement of Brno University of Technology, Institute of Aerospace Engineering in the AG52. The considered test case is based on the RAE2822 aerofoil constrained multipoint optimization problem. The overall problem setup is given along with selected surrogate modelling and optimization techniques. The presented approach featuring artificial neural networks is able to produce highly reliable metamodels with cutting-edge performance as documented by the AG52 performance metrics comparison. The work described within this paper has been presented at EUROGEN 2015 conference in Glasgow.
Keywords in Czech
náhradní model, metamodel, vícekriteriální optimalizace, vícebodová optimalizace, Garteur AG52, RAE2822
Keywords in English
surrogate modelling, metamodel, multi-objective optimization, multi-point, Garteur AG52, RAE2822
RIV year
2015
Released
30.12.2015
Publisher
ALV - Association of the Aerospace Manufacturers
Location
Prague
ISSN
1211-877X
Volume
2015
Number
3
Pages from–to
5–10
Pages count
6
BIBTEX
@article{BUT120867,
author="Petr {Dvořák},
title="Surrogate-based Shape Optimization in Aerodynamic Applications",
year="2015",
volume="2015",
number="3",
month="December",
pages="5--10",
publisher="ALV - Association of the Aerospace Manufacturers",
address="Prague",
issn="1211-877X"
}