Aerodynamics-guided machine learning for design optimization of electric vehicles

Abstract The transition to electric vehicles is driving a fundamental shift in the automobile design process. Changes in constraints afforded by the absence of a combustion engine create new opportunities for modifying vehicle geometries. Current approaches to optimizing vehicle aerodynamics require...

Full description

Saved in:
Bibliographic Details
Main Authors: Jonathan Tran, Kai Fukami, Kenta Inada, Daisuke Umehara, Yoshimichi Ono, Kenta Ogawa, Kunihiko Taira
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:Communications Engineering
Online Access:https://doi.org/10.1038/s44172-024-00322-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850162536216788992
author Jonathan Tran
Kai Fukami
Kenta Inada
Daisuke Umehara
Yoshimichi Ono
Kenta Ogawa
Kunihiko Taira
author_facet Jonathan Tran
Kai Fukami
Kenta Inada
Daisuke Umehara
Yoshimichi Ono
Kenta Ogawa
Kunihiko Taira
author_sort Jonathan Tran
collection DOAJ
description Abstract The transition to electric vehicles is driving a fundamental shift in the automobile design process. Changes in constraints afforded by the absence of a combustion engine create new opportunities for modifying vehicle geometries. Current approaches to optimizing vehicle aerodynamics require a vast amount of computational studies and physical experiments, which are expensive when performing parameter sweeps over conceivable geometric configurations, suggesting the need for more efficient surrogate models to assist analysis. Here we analyze a dataset of industry-quality automobile geometries with their associated aerodynamic performance obtained from experimentally validated, high-fidelity large-eddy simulations. We show that a relationship between these geometries and their respective aerodynamics can be extracted in a low-dimensional manner by leveraging a nonlinear autoencoder which is simultaneously trained to estimate the drag coefficient from the latent variables. We perform aerodynamic design optimization of vehicle designs by making use of the learned aerodynamic relationship in the low-order space obtained by the model. We demonstrate that the aerodynamic trends for the geometries produced from the optimization process show agreement with validation simulations. The findings of this work demonstrate the application of data-driven approaches to the analysis and design of vehicles in a production environment.
format Article
id doaj-art-64cc5835b71945f0bc9162c48614df0d
institution OA Journals
issn 2731-3395
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series Communications Engineering
spelling doaj-art-64cc5835b71945f0bc9162c48614df0d2025-08-20T02:22:33ZengNature PortfolioCommunications Engineering2731-33952024-11-01311910.1038/s44172-024-00322-0Aerodynamics-guided machine learning for design optimization of electric vehiclesJonathan Tran0Kai Fukami1Kenta Inada2Daisuke Umehara3Yoshimichi Ono4Kenta Ogawa5Kunihiko Taira6Department of Mechanical and Aerospace Engineering, University of CaliforniaDepartment of Mechanical and Aerospace Engineering, University of CaliforniaBEV Automobile Development Unit, Honda Motor Co., Ltd.BEV Automobile Development Unit, Honda Motor Co., Ltd.BEV Automobile Development Unit, Honda Motor Co., Ltd.BEV Automobile Development Unit, Honda Motor Co., Ltd.Department of Mechanical and Aerospace Engineering, University of CaliforniaAbstract The transition to electric vehicles is driving a fundamental shift in the automobile design process. Changes in constraints afforded by the absence of a combustion engine create new opportunities for modifying vehicle geometries. Current approaches to optimizing vehicle aerodynamics require a vast amount of computational studies and physical experiments, which are expensive when performing parameter sweeps over conceivable geometric configurations, suggesting the need for more efficient surrogate models to assist analysis. Here we analyze a dataset of industry-quality automobile geometries with their associated aerodynamic performance obtained from experimentally validated, high-fidelity large-eddy simulations. We show that a relationship between these geometries and their respective aerodynamics can be extracted in a low-dimensional manner by leveraging a nonlinear autoencoder which is simultaneously trained to estimate the drag coefficient from the latent variables. We perform aerodynamic design optimization of vehicle designs by making use of the learned aerodynamic relationship in the low-order space obtained by the model. We demonstrate that the aerodynamic trends for the geometries produced from the optimization process show agreement with validation simulations. The findings of this work demonstrate the application of data-driven approaches to the analysis and design of vehicles in a production environment.https://doi.org/10.1038/s44172-024-00322-0
spellingShingle Jonathan Tran
Kai Fukami
Kenta Inada
Daisuke Umehara
Yoshimichi Ono
Kenta Ogawa
Kunihiko Taira
Aerodynamics-guided machine learning for design optimization of electric vehicles
Communications Engineering
title Aerodynamics-guided machine learning for design optimization of electric vehicles
title_full Aerodynamics-guided machine learning for design optimization of electric vehicles
title_fullStr Aerodynamics-guided machine learning for design optimization of electric vehicles
title_full_unstemmed Aerodynamics-guided machine learning for design optimization of electric vehicles
title_short Aerodynamics-guided machine learning for design optimization of electric vehicles
title_sort aerodynamics guided machine learning for design optimization of electric vehicles
url https://doi.org/10.1038/s44172-024-00322-0
work_keys_str_mv AT jonathantran aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT kaifukami aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT kentainada aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT daisukeumehara aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT yoshimichiono aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT kentaogawa aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles
AT kunihikotaira aerodynamicsguidedmachinelearningfordesignoptimizationofelectricvehicles