基于多精度深度神经网络的汽车气动外形优化设计方法

邬晓敬; 高然; 马龙

doi:10.7638/kqdlxxb-2023.0164

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

基于多精度深度神经网络的汽车气动外形优化设计方法

研究专栏——汽车空气动力学 | 更新时间：2026-07-27

- 基于多精度深度神经网络的汽车气动外形优化设计方法
- Optimization design method of automobile aerodynamic shape based on multi-fidelity deep neural network
- 空气动力学学报 2024年42卷第7期页码：103-111
- 作者机构：
  
  西北工业大学民航学院,西安,710072
- 作者简介：
  
  邬晓敬*（1989—），男，河南信阳人，副教授，研究方向：气动外形优化设计. E-mail：wuxj@nwpu.edu.cn
- 基金信息：
- DOI：10.7638/kqdlxxb-2023.0164
  中图分类号： U46
- 纸质出版：2024
- 稿件说明：
移动端阅览
邬晓敬, 高然, 马龙. 基于多精度深度神经网络的汽车气动外形优化设计方法[J]. 空气动力学学报, 2024, 42(7): 103−111.

WU Xiaojing, GAO Ran, MA Long. Optimization design method of automobile aerodynamic shape based on multi-fidelity deep neural network[J]. 2024, 42(7): 103-111.
邬晓敬, 高然, 马龙. 基于多精度深度神经网络的汽车气动外形优化设计方法[J]. 空气动力学学报, 2024, 42(7): 103−111. DOI： 10.7638/kqdlxxb-2023.0164

WU Xiaojing, GAO Ran, MA Long. Optimization design method of automobile aerodynamic shape based on multi-fidelity deep neural network[J]. 2024, 42(7): 103-111. DOI： 10.7638/kqdlxxb-2023.0164.

摘要

在汽车气动外形优化设计中，往往需要大量的高精度CFD数据作为支撑。然而，高精度CFD数据获取难度大、成本高。为了缓解汽车气动优化设计中气动特性评估精度和效率之间的矛盾，根据迁移学习与数据融合的思想，提出了一种基于多精度深度神经网络（multi-fidelity deep neural network

MFDNN）的汽车外形优化设计方法，以减少优化设计中所需的高精度数据个数，从而有效提升优化速度、降低优化成本。将所发展的优化方法应用于快背式MIRA标准模型减阻优化设计中，优化结果表明，该方法能够充分融合不同精度数据所蕴含的知识，加速气动外形优化进程，提升优化效率。以收敛用时作为评价指标，在取得相近或更优优化结果的前提下，基于多精度神经网络的优化框架的收敛速度是基于单精度神经网络的离线优化框架的5.85倍，是基于单精度神经网络的在线优化框架的2.81倍。

Abstract

The optimization process of automobile aerodynamic design that requires a vast amount of CFD data

which are both expensive and time-consuming to acquire

is arduous and costly. To get out of this predicament—the conflict between accuracy and efficiency—we devise an innovative automobile shape optimization technique relying on the Multi-fidelity deep neural network (MFDNN) aerodynamic model based on transfer learning and data fusion. Applying the developed optimization method to the drag reduction optimization design of the fast-back MIRA standard model demonstrates that the method can fully integrate the knowledge contained in different fidelity data

accelerate the aerodynamic shape optimization process

and improve the optimization efficiency. Specifically

the convergence speed of the optimization framework based on multi-precision neural network is 5.85 times that of the offline optimization framework based on single-precision neural network and 2.81 times that of the online optimization framework based on single-precision neural network.

关键词

Keywords

references

FORRESTER A I J, KEANE A J . Recent advances in surrogate-based optimization[J ] . Progress in Aerospace Sciences, 2009 , 45 ( 1-3 ): 50 − 79 .

ZHANG Y C, XUE X D, DING W, et al . Automatic shape optimization of hatchback to reduce aerodynamic drag[J ] . Journal of Tongji University (Natural Science), 2016 , 44 ( 11 ): 1771 − 1775, 1795 (in Chinese).

LAI C G, DUAN M H, CHEN X X, et al. Multi-objective optimization of vehicle aerodynamic shape parameters based on an intelligent algorithm[J]. Journal of Beijing University of Technology, 2017, 43(9): 1320−1327(in Chinese).

ZHANG Y C, XUE X D, DING W, et al. Automatic shape optimization of hatchback to reduce aerodynamic drag[J]. Journal of Tongji University (Natural Science), 2016, 44(11): 1771−1775, 1795(in Chinese).

HAN Z H, ZIMMERMANN, GÖRTZ S. Alternative cokriging method for variable-fidelity surrogate modeling[J]. AIAA Journal, 2012, 50(5): 1205−1210.

HAN Z H, GÖRTZ S. Hierarchical Kriging model for variable-fidelity surrogate modeling[J]. AIAA Journal, 2012, 50(9): 1885−1896.

BONFIGLIO L, PERDIKARIS P, BRIZZOLARA S, et al. Multi-fidelity optimization of super-cavitating hydrofoils[J]. Computer Methods in Applied Mechanics and Engineering, 2018, 332: 63−85.

ZHENG H Y, XIE F F, JI T W, et al . Multifidelity kinematic parameter optimization o f a flapping airfoil[J ] . Physical Review E, 2020 , 101 ( 1-1 ): 013107 .

DING F, KAREEM A. A multi-fidelity shape optimization via surrogate modeling for civil structures[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2018, 178: 49−56.

ZHANG Y, HAN Z H, ZHANG K S. Variable-fidelity expected improvement method for efficient global optimization of expensive functions[J]. Structural and Multidisciplinary Optimization, 2018, 58(4): 1431−1451.

ZHANG X S, XIE F F, JI T W, et al. Multi-fidelity deep neural network surrogate model for aerodynamic shape optimization[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 373: 113485.

HAN Z H, XU C Z, ZHANG L, et al. Efficient aerodynamic shape optimization using variable-fidelity surrogate models and multilevel computational grids[J]. Chinese Journal of Aeronautics, 2020, 33(1): 31−47.

KOU J Q, NING C J, ZHANG W W. Transfer learning for flow reconstruction based on multifidelity data[J]. AIAA Journal, 2022, 60(10): 5821−5842.

HOFFMAN J, MARTINDALE B, ARNETTE S, et al. Effect of test section configuration on aerodynamic drag measurements[J]. SAE Transactions, 2001, 110: 680−694.

GU Z Q, WANG S, QIU J, et al. Wind tunnel tests of MIRA model group for study of vehicle's rear shape[J]. Science and Technology Review, 2011, 29(8): 67−71(in Chinese).

MENG X H, KARNIADAKIS G E. A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems[J]. Journal of Computational Physics, 2020, 401: 109020.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向三维飞行器的气动性能智能预测模型及迁移学习框架研究

基于迁移学习的点云机翼气动性能预测

面向民机标模的宽域高雷诺数小样本气动建模方法研究

基于 XGBoost 的多源气动数据融合建模

基于物理信息神经网络的气动数据融合方法