All Issue

2024 Vol.37, Issue 5 Preview Page

Research Paper

Development of a Reduced Order Model using a Deep Learning-based Manifold-Augmented Approach

매니폴드 데이터 증강기법 기반의 딥러닝 방법론을 적용한 축소 모델 개발

Seongwoo Cheon1
,
Hyejin Kim2
,
Seokhee Ryu34
,
Haeseong Cho5*
,
Hakjin Lee6
천 성우1
,
김 혜진2
,
류 석희34
,
조 해성5*
,
이 학진6
1Graduate Student, Department of Aeropsace Engineering, Jeonbuk National University, Jeonju, 54896, Korea
2Graduate Student, Department of Aeropsace Engineering, Jeonbuk National University, Jeonju, 54896, Korea
3Researcher, Department of Mechanical Aerospace Engineering, Gyeongsang National University, Jinju, 52828, Korea
4Researcher, Quality Planning Team, Defense Agency of Technology and Quality, Jinju, 52851, Korea
5Asosociate Professor, Department of Aeropsace Engineering, Jeonbuk National University, Jeonju, 54896, Korea
6Associate Professor, Department of Mechanical Aerospace Engineering, Gyeongsang National University, Jinju, 52828, Korea
1전북대학교 항공우주공학과 박사과정
2전북대학교 항공우주공학과 석박사통합과정
3경상국립대학교 기계항공공학부 연구원
4국방기술품질원 품질기획실 연구원
5전북대학교 항공우주공학과 부교수
6경상국립대학교 기계항공공학부 부교수
*Corresponding Author
31 October 2024. pp. 337-344
PDF XML Citation
Abstract

This study presents a deep learning-based framework to predict the aerodynamic performance of low Reynolds number airfoils. The framework employs a convolutional neural network (CNN) combined with a variational autoencoder (VAE) to efficiently handle large datasets. Moreover, the signed distance function is used as the network input to represent the airfoil configuration in the image data and parameterize the CNN. A novel generative model based on projection-based manifold learning is proposed to overcome the data mining limitation of computational fluid dynamics which may incur significant computational costs. The interpolation and extrapolation accuracy of the proposed framework is evaluated using the NACA 4-digit airfoil configuration.The results show improved accuracy via data augmentation performed by the proposed generative model.

Keywords
manifold learning
model-order reduction
deep learning
data augmentation

본 논문에서는 저 레이놀즈 수 영역에서 에어포일의 공기역학적 성능을 예측하기 위한 딥러닝 기반의 축소 모델을 제시하였다. 딥러닝 기반 축소 모델에서 CFD 해석 결과의 높은 차원의 데이터를 효율적으로 다루기 위해 변이형 오토인코더를 결합한 합성곱 신경망을 적용하였다. 부호화 거리 함수를 통해 에어포일의 형상과 유동 조건을 이미지 데이터화 하고, 이에 대해 합성곱 신경망을 매개변수화 하였다. 또한, 전산유체역학 해석의 계산 비용으로 인한 부족한 훈련 데이터를 극복하기 위해 투영 기반의 비선형 매니폴드 데이터 증강기법을 개발하였다. NACA 4계열 에어포일은 해석 예제로 고려하여 제안하는 프레임워크의 내삽과 외삽 정확도를 평가하였으며 매니폴드 데이터 증강기법을 적용하여 프레임워크의 정확도 향상을 확인하였다.

키워드
매니폴드 러닝
모델 차수 축소
딥러닝
연산유체역학
데이터 증강
References
1

Balla, K., Sevilla, R., Hassan, O., Morgan, K. (2021) An Application of Neural Networks to the Prediction of Aerodynamic Coefficients of Aerofoils and Wings, Appl. Math. Model., 96, pp.456~479.

10.1016/j.apm.2021.03.019
2

Bhatnagar, S., Afshar, Y., Pan, S., Duraisamy K., Kaushik, S. (2019) Prediction of Aerodynamic Flow Fields using Convolutional Neural Networks, Comput. Mech., 64, pp.525~545.

10.1007/s00466-019-01740-0
3

Chadebec, C., Allassonnière, S. (2021) Data Augmentation with Variational Autoencoders and Manifold Sampling, In Deep Generative Models, and Data Augmentation, Labelling, and Imperfections: First Workshop, DGM4MICCAI 2021, and First Workshop, DALI 2021, Held in Conjunction with MICCAI 2021, Proceedings 1, pp.184~192.

10.1007/978-3-030-88210-5_17
4

Chadebec, C., Allassonnière, S. (2022) A Geometric Perspective on Variational Autoencoders, Adv. Neural Inf. Proc. Syst., 35, pp.19618~19630.

5

Geelen, R., Wright, S., Willcox, K. (2023) Operator Inference for Non-Intrusive Model Reduction with Quadratic Manifolds, Comput. Methods App. Mech. & Eng., 403, p.115717.

10.1016/j.cma.2022.115717
6

Jia, X., Li, C., Ji, W., Gong, C. (2022) A Hybrid Reduced-Order Model Combing Deep Learning for Unsteady Flow, Phys. Fluids, 34(9), p.097112.

10.1063/5.0104848
7

Lee, S., Jang, K., Cho, H., Kim, H., Shin, S. (2021) Parametric Non-Intrusive Model order Reduction for Flow-Fields using Unsupervised Machine Learning, Comput. Methods Appl. Mech. & Eng., 384, p.113999.

10.1016/j.cma.2021.113999
8

Mumuni, A., Mumuni, F. (2022) Data Augmentation: A Comprehensive Survey of Modern Approaches, Array, 16, p.100258.

10.1016/j.array.2022.100258
9

Salimans, T., Goodfellow, I., Zaremba, W., Cheung, V., Radford, A., Chen, X. (2016) Improved Techniques for Training GANs, Adv. Neural Inf. Proc. Syst., 29.

10

Spalart, P.R., Venkatakrishnan, V. (2016) On the Role and Challenges of CFD in the Aerospace Industry, The Aeronaut. J., 120(1223), pp.209~232.

10.1017/aer.2015.10
11

Xia, B., Sun, D.-W. (2002) Applications of Computational Fluid Dynamics (CFD) in the Food Industry: A Review, Comput. & Electron. Agric., 34(1-3), pp.5~24.

10.1016/S0168-1699(01)00177-6
Information
  • Publisher :Computational Structural Engineering Institute of Korea
  • Publisher(Ko) :한국전산구조공학회
  • Journal Title :Journal of the Computational Structural Engineering Institute of Korea
  • Journal Title(Ko) :한국전산구조공학회 논문집
  • Volume : 37
  • No :5
  • Pages :337-344
  • Received Date : 2024-07-10
  • Revised Date : 2024-08-26
  • Accepted Date : 2024-08-27
  • DOI :https://doi.org/10.7734/COSEIK.2024.37.5.337
Journal Informaiton Journal of the Computational Structural Engineering Institute of Korea Journal of the Computational Structural Engineering Institute of Korea
Online Submission
  • NRF
  • KOFST
  • crossref crossmark
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close