Title: Enhanced data efficiency using deep neural networks and Gaussian processes for aerodynamic design optimization

Abstract

Adjoint-based optimization methods are attractive for aerodynamic shape design primarily due to their computational costs being independent of the dimensionality of the input space and their ability to generate high-fidelity gradients that can then be used in a gradient-based optimizer. This makes them very well suited for high-fidelity simulation based aerodynamic shape optimization of highly parametrized geometries such as aircraft wings. However, the development of adjoint-based solvers involve careful mathematical treatment and their implementation require detailed software development. Furthermore, they can become prohibitively expensive when multiple optimization problems are being solved, each requiring multiple restarts to circumvent local optima. In this work, we propose a machine learning enabled, surrogate-based framework that replaces the expensive adjoint solver, without compromising on predicting predictive accuracy. Specifically, we first train a deep neural network (DNN) from training data generated from evaluating the high-fidelity simulation model on a model-agnostic design of experiments on the geometry shape parameters. The optimum shape may then be computed by using a gradient-based optimizer coupled with the trained DNN. Subsequently, we also perform a gradient-free Bayesian optimization, where the trained DNN is used as the prior mean. We observe that the latter framework (DNN-BO) improves upon the DNN-only based optimizationmore » strategy for the same computational cost. Overall, this framework predicts the true optimum with very high accuracy, while requiring far fewer high-fidelity function calls compared to the adjoint-based method. Furthermore, we show that multiple optimization problems can be solved with the same machine learning model with high accuracy, to amortize the offline costs associated with constructing our models. Our methodology finds applications in the early stages of aerospace design.« less

Authors:

Renganathan, S. Ashwin  ^[1];

• Search DOE PAGES for author "Renganathan, S. Ashwin"
• Search DOE PAGES for ORCID "0000-0001-6948-6932"

Maulik, Romit ^[1]; Ahuja, Jai  ^[2]

• Search DOE PAGES for author "Ahuja, Jai"
• Search DOE PAGES for ORCID "0000-0002-4370-0416"

---

+ Show Author Affiliations

1. Argonne National Lab. (ANL), Argonne, IL (United States)
2. Georgia Inst. of Technology, Atlanta, GA (United States)

Publication Date:

Thu Apr 01 00:00:00 EDT 2021

Research Org.:

Argonne National Laboratory (ANL), Argonne, IL (United States)

Sponsoring Org.:

USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)

OSTI Identifier:

1798159

Grant/Contract Number:  

AC02-06CH11357

Resource Type:

Accepted Manuscript

Journal Name:

Aerospace Science and Technology

Additional Journal Information:

Journal Volume: 111; Journal ID: ISSN 1270-9638

Publisher:

Elsevier

Country of Publication:

United States

Language:

English

Subject:

42 ENGINEERING; Adjoint method; Aerodynamic design optimization; Bayesian optimization; Gaussian processes; Machine learning