基于知识迁移的高效涡轮叶栅智能气动优化方法

doi:10.3901/JME.2025.04.333

机械工程学报 ›› 2025, Vol. 61 ›› Issue (4): 333-343.doi: 10.3901/JME.2025.04.333

• 交叉与前沿 • 上一篇

扫码分享

基于知识迁移的高效涡轮叶栅智能气动优化方法

李存晰, 宋立明, 郭振东, 李军

西安交通大学叶轮机械研究所西安 710049

收稿日期:2024-04-07 修回日期:2024-10-28 发布日期:2025-04-14
作者简介:李存晰，男，2002年出生。主要研究方向为智能气动形状参数化与优化方法。E-mail：licunxi@stu.xjtu.edu.cn
郭振东(通信作者)，男，1990年出生，博士，副教授。主要研究方向为高效精细气动设计优化与知识挖掘、智能多学科优化和基于深度学习的流场预测。E-mail：guozhendong@xjtu.edu.cn
基金资助:
国家自然科学基金(52306048)和国家科技重大专项(2019-II-0008-0028)资助项目。

Knowledge Transfer Based Aerodynamic Optimization Method for Efficient Turbine Cascade Design

LI Cunxi, SONG Liming, GUO Zhendong, LI Jun

Institute of Turbomachinery, Xi'an Jiaotong University, Xi'an 710049

Received:2024-04-07 Revised:2024-10-28 Published:2025-04-14

摘要/Abstract

摘要： 为实现涡轮叶栅气动形状的高效优化，基于人工智能领域知识迁移理念，开展基于知识迁移的高效涡轮叶栅智能气动优化方法研究。首先，搭建叶型变分自动编码器模型，利用其解码器实现气动形状参数化；其次，为提高已完成任务样本重参数化精度，提出基于样本加权的模型重训练策略，并设计出基于梯度信息的叶型重参数化算法，将已完成任务样本编码至目标任务空间；然后，实现基于多保真度模型的贝叶斯迁移优化算法，以有效利用迁移源任务信息加速目标任务优化进程；最后，搭建智能涡轮叶栅气动迁移优化框架。通过在GE-E³低压涡轮算例中开展叶型设计优化表明，在计算成本相同的情况，知识迁移策略所获得的最优解相对参考设计总压损失减少0.66%，相对无知识迁移策略所获得的最优解总压损失减少0.18%；而在获得相似性能优化解的情况下，知识迁移策略所需计算成本相对传统无知识迁移方法降低50%以上。由此验证了所提出框架的性能优势。

关键词: 知识迁移, 贝叶斯优化, 多保真度代理模型, 涡轮气动形状优化

Abstract: For the efficient aerodynamic shape optimization(ASO) of turbine blades, inspired by the concept of knowledge transfer from artificial intelligence, research on knowledge transfer based intelligent aerodynamic shape optimization methods for turbine blades is conducted. Firstly, the variational autoencoder(VAE) model is built, with its decoder to realize the intelligent parameterization of turbine blades. Secondly, for the improvement of the reparameterization accuracy of samples from related completed tasks, a sample-weighted model retraining strategy is proposed for encoding samples to a common design space. Besides, a gradient-based reparameterization algorithm is also designed. Moreover, the Bayesian transfer optimization algorithm is constructed based on multi-fidelity surrogates to take full advantage of transfer source samples and accelerate the optimization progress of target tasks. Eventually, an intelligent transfer ASO framework for turbine blades is established. Transfer optimization for low pressure turbine blade of GE-E³ is carried out using the proposed framework. With the same computational cost, the total pressure loss coefficient of the optimal blade is reduced by 0.66% compared to benchmark design, and 0.18% lower than legacy approach without knowledge transfer strategy. With reaching the same level of optimized performance, the computational cost of the optimization with proposed framework can be reduced by at least 50% compared to legacy approach. Thereby, the effectiveness of proposed knowledge transfer accelerated ASO framework for turbine blades is well demonstrated.

Key words: knowledge transfer, Bayesian optimization, multi-fidelity surrogate, turbine aerodynamic shape optimization

中图分类号:

TK14

李存晰, 宋立明, 郭振东, 李军. 基于知识迁移的高效涡轮叶栅智能气动优化方法[J]. 机械工程学报, 2025, 61(4): 333-343.

LI Cunxi, SONG Liming, GUO Zhendong, LI Jun. Knowledge Transfer Based Aerodynamic Optimization Method for Efficient Turbine Cascade Design[J]. Journal of Mechanical Engineering, 2025, 61(4): 333-343.

参考文献

[1] SHAN Songqing，WANG G G. Survey of modeling and optimization strategies to solve high-dimensional design problems with computationally-expensive black-box functions[J]. Structural and Multidisciplinary Optimization，2010，41(2):219-241.
[2] LI Jichao，ZHANG Mengqi. Data-based approach for wing shape design optimization[J]. Aerospace Science and Technology，2021，112:106639.
[3] KORAKIANITIS T. Hierarchical development of three direct-design methods for two-dimensional axial-turbomachinery cascades[J]. Journal of Turbomachinery，1993，115(2):314-324.
[4] GOODFELLOW I，POUGET-ABADIE J，MIRZA M，et al. Generative adversarial networks[DB/OL]. arXiv，2014[2014-06-10]. https://arxiv.org/abs/1406.2661.
[5] CHEN Wei，CHIU K，FUGE M. Aerodynamic design optimization and shape exploration using generative adversarial networks[C]//American Institute of Aeronautics and Astronautics. AIAA Scitech 2019 Forum，January 7-11，San Diego，California. Reston，Virginia:AIAA，2019:2351.
[6] GUPTA A，ONG Y S，FENG Liang. Insights on transfer optimization:Because experience is the best teacher[J]. IEEE Transactions on Emerging Topics in Computational Intelligence，2018，2(1):51-64.
[7] PAN S J，YANG Qiang. A survey on transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering，2010，22(10):1345-1359.
[8] TAN K C，FENG Liang，JIANG Min. Evolutionary transfer optimization—A new frontier in evolutionary computation research[J]. IEEE Computational Intelligence Magazine，2021，16(1):22-33.
[9] MIN A T W，SAGARNA R，GUPTA A，et al. Knowledge transfer through machine learning in aircraft design[J]. IEEE Computational Intelligence Magazine，2017，12(4):48-60.
[10] TAN A W，SAGARNA R，GUPTA A，et al. Coping with data scarcity in aircraft engine design[C]//American Institute of Aeronautics and Astronautics. 18th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference，June 5-9，2017，Denver，Colorado. Reston，Virginia:AIAA，2017. https://doi.org/10.2514/6.2017-4434.
[11] WANG Qineng，SONG Liming，GUO Zhendong，et al. Transfer optimization in accelerating the design of turbomachinery cascades[C/OL]//American Society of Mechanical Engineers. Turbo Expo:Power for Land，Sea，and Air，Volume 2D:Turbomachinery，September 21-25，2020. New York:ASME，2020. https://doi.org/10.1115/GT2020-16321.
[12] GUO Zhendong，SUN Wei，SONG Liming，et al. Generative transfer optimization for aerodynamic design[C]//Global Power and Propulsion Society. GPPS Xi’an21，October 18-20，2021，Xi’an，Shaanxi，China. GPPS，2021:225.
[13] GUO Zhendong，LIU Haitao，ONG Y S，et al. Generative multiform Bayesian optimization[J]. IEEE Transactions on Cybernetics，2023，53(7):4347-4360.
[14] KINGMA D P，WELLING M. Auto-encoding variational Bayes[DB/OL]. arXiv，2013[2022-12-10]. https://arxiv.org/abs/1312.6114.
[15] VALENZUELA DEL RÍO J E，MAVRIS D. Gaussian process surrogate model for levering similar trends across concepts[J]. AIAA Journal，2015，53(4):1002-1015.
[16] YANG Yunjia，LI Runze，ZHANG Yufei，et al. Flowfield prediction of airfoil off-design conditions based on a modified variational autoencoder[J]. AIAA Journal，2022，60(10):5805-5820.
[17] TRIPP A，DAXBURGER E，HERNANDEZ-LOBATO J. Sample-efficient optimization in the latent space of deep generative models via weighted retraining[DB/OL]. arXiv，2006[2020-10-25]. https://arxiv.org/abs/2006.09191.
[18] SHAHRIARI B，SWERSKY K，WANG Ziyu，et al. Taking the human out of the loop:A review of Bayesian optimization[J]. Proceedings of the IEEE，2015，104(1):148-175.
[19] RASMUSSEN C E，WILLIAMS C K I. Gaussian process for machine learning[M/OL] Cambridge，Massachusetts:The MIT Press，2006[2022-04-01]. https://gaussianprocess.org/gpml/.
[20] GUO Zhendong，WANG Qineng，SONG Liming，et al. Parallel multi-fidelity expected improvement method for efficient global optimization[J]. Structural and Multidisciplinary Optimization，2021，64(3):1457-1468.
[21] STEIN M. Large sample properties of simulations using latin hypercube sampling[J]. Technometrics，1987，29(2):143-151.
[22] 航空航天工业部高效节能发动机文集编委会. 高效节能发动机文集第五分册[M]. 北京:航空工业出版社，1991. Editorial Committee of the Proceedings of the Energy Efficient Engine，Ministry of Aviation and Aerospace Industry. Proceedings of the energy efficient engine，volume 5[M]. Beijing:Aviation Industry Press，1991.

基于知识迁移的高效涡轮叶栅智能气动优化方法

Knowledge Transfer Based Aerodynamic Optimization Method for Efficient Turbine Cascade Design

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

本文评价

[1]	林亮行, 马国政, 孙建芳, 韩翠红, 雍青松, 苏峰华, 王海斗. 基于VMD-EEMD-LSTM的涂层型关节轴承剩余使用寿命预测方法[J]. 机械工程学报, 2023, 59(9): 125-136.
[2]	吴西涛, 魏超, 翟建坤, 苑士华. 考虑横摆稳定性的无人车轨迹跟踪控制优化研究[J]. 机械工程学报, 2022, 58(6): 130-142.
[3]	贾寒冰, 刘鹏, 张雷, 王震坡. 基于规则与机器学习融合的换道决策建模方法研究[J]. 机械工程学报, 2022, 58(4): 212-221.