在对翼型Bezier-PARSEC参数化软件FanOpt的应用方式改进后建立数据集,以翼型升阻比特性为目标,分别利用支持向量回归（SVR）、决策树、随机森林回归、全连接神经网络、一维卷积神经网络等机器学习模型进行拟合,训练模型比较拟合精度。结果表明,全连接神经网络、一维卷积神经网络作为替代计算模型在测试集上对升阻比的预测准确率可达97.86%,但相比于一维卷积神经网络,全连接神经网络在处理这种结构不复杂的数据集时更有优势。
Abstract
After the improvement of the application mode of airfoil Bezier-PARSEC parameterization software FanOpt, the data set is established. Taking the Lift-to-Drag Ratio characteristics of airfoil as the target, the machine learning models such as support vector regression （SVR）, decision tree, random forest regression, fully connected neural network and one-dimensional convolution neural network are used to fit the data, and the fitting accuracy of the training models is compared. The results show that the prediction accuracy of lift to drag ratio can reach 97.86% by using fully connected neural network and one-dimensional convolutional neural network as surrogate calculation models in the test set. However, compared with one-dimensional convolutional neural network, fully connected neural network has more advantages in dealing with this kind of data set with uncomplicated structure.
关键词
潮流能 /
叶片 /
机器学习 /
水轮机翼型 /
参数化 /
神经网络
Key words
tidal energy /
blades /
machine learning /
hydrofoils /
parameterization /
neural networks
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 黄继鸿, 苏红莲, 赵新华. 基于BP神经网络的翼型空气动力系数预测[J]. 航空工程进展, 2010, 1(1): 36-39.
HUANG J H, SU H L, ZHAO X H.Aerody-namic coefficient prediction of airfoil using BP neural network[J]. Advances in aeronautical science and engineering, 2010, 1(1): 36-39.
[2] SANTOS M, MATTOS B, GIRARDI R.Aerodynamic coefficient prediction of airfoils using neural networks[C]//46th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, 2008.
[3] YILMAZ E, GERMAN B.A convolutional neural network approach to training predictors for airfoil performance[C]//18th AIAA/Issmo Multidisciplinary Analysis & Optimization Conference, Denver, CO, USA, 2017.
[4] SEKAR V, ZHANG M Q, SHU C, et al.Inverse design of airfoil using a deep convolutional neural network[J]. AIAA journal, 2019, 57(2): 1-11.
[5] CHEN H, HE L, QIAN W, et al.Multiple aerodynamic coefficient prediction of airfoils using a convolutional neural network[J]. Symmetry, 2020, 12(4): 544-558.
[6] 王邦祥, 陆金桂, 王京涛. 神经网络在风力机翼型气动性能优化中的应用[J]. 机械设计与制造, 2020, 349(3): 243-247.
WANG B X, LU J G, WANG J T.Application of neural network to aerodynamic performance optimization of wind turbine[J]. Machinery design and manufacture, 2020, 349(3): 243-247.
[7] 朱国俊, 冯建军, 郭鹏程, 等. 基于径向基神经网络-遗传算法的海流能水轮机叶片翼型优化[J]. 农业工程学报, 2014, 30(8): 65-73.
ZHU G J, FENG J J, GUO P C, et al.Optimization of hydrofoil for marine current turbine based on radial basis function neural network and genetic algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering, 2014,30(8): 65-73.
[8] PATRI A, PATNAIK Y.Random forest and stochastic gradient tree boosting based approach for the prediction of airfoil self-noise[J]. Procedia computer science, 2015, 46: 109-121.
[9] 孙勤. 水平轴风力机翼型优化设计方法的研究[D]. 湘潭: 湘潭大学, 2013.
SUN Q.Research on optimal design of airfoil for horizontal-axis wind turbines[D]. Xiangtan: Xiangtan University, 2013.
[10] 朱尧华, 袁寿其, 张金凤, 等. 水轮机叶片翼型参数化研究概述[J]. 中国农村水利水电, 2016(1): 129-132.
ZHU Y H, YUAN S Q, ZHANG J F, et al.Research status for parameterization hydro-turbine blade airfoil[J]. China rural water and hydropower, 2016(1): 129-132.
[11] HICKS R M, HENNE P A.Wing design by numerical optimization[J]. Journal of aircraft, 1978, 15(7): 407-412.
[12] SOBIECZKY H.Parametric airfoils and wings, notes on numerical fluid mechanics[M]. Wiesbaden: Vieweg+Teubner Verlag, 1998: 71-88.
[13] DERKSEN R W, ROGALSKY T.Bezier-parsec: an optimized aerofoil parameterization for design[J]. Advances in engineering software, 2010, 41(7-8): 923-930.
[14] ROGALSKY T P.Acceleration of differential evolution for aerodynamic design[D]. Manitoba: University of Manitoba, 2004.
[15] 周志华, 王珏. 机器学习及其应用2009[M]. 北京: 清华大学出版社, 2009: 97-101.
ZHOU Z H, WANG J.Machine learning and its application 2009[M]. Beijing: Tsinghua University Press, 2009: 97-101.
[16] IOFFE S, SZEGEDY C.Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 2015.
基金
国家重点研发计划（2018YFB1501903）; 山东省重点研发计划（2019GGX103012）
PDF(1886 KB)
