基于深度学习的翼型反设计方法

何磊; 钱炜祺; 刘滔

doi:10.13224/j.cnki.jasp.2020.09.013

基于深度学习的翼型反设计方法

doi: 10.13224/j.cnki.jasp.2020.09.013

基金项目: 国家自然科学基金（11802325）; 中国空气动力研究与发展中心基础和前沿技术研究基金（FFTRF20172015）

计量
- 文章访问数: 224
- HTML浏览量: 4
- PDF量: 269
- 被引次数: 0
出版历程
- 收稿日期: 2020-03-15
- 刊出日期: 2020-09-28

Inverse design method of airfoil based on deep learning

Computational Aerodynamics Institute,China Aerodynamics Research and Development Center,Mianyang Sichuan 621000,China

摘要

摘要: 建立了一种基于深度学习的翼型反设计方法,将翼型曲线及其对应的压力分布图像作为训练学习对象,建立其内在联系的模型,实现通过卷积神经网络提取压力分布图像的特征,计算获得翼型曲线。该方法直接将压力分布图像作为模型输入,更加直观简洁,同时避免了传统方法中耗时的数值计算过程。模型测试中,6 000组压力分布图像和翼型曲线用于模型训练,另外561组用于模型验证,验证耗时仅67 s,预测的翼型曲线与CFD计算结果的平均相对误差为055%。对比实验中,通过对压力分布曲线添加噪声、改变输出层尺寸等方式,进一步验证和分析了预测模型性能。结果表明该翼型反设计方法具有较高预测精度和较强鲁棒性,能在保证精度的情况下降低计算时间,提高设计效率。
- 反设计方法 /
- 翼型曲线 /
- 压力分布 /
- 卷积神经网络 /
- 深度学习 /
- 预测模型
Abstract: An inverse design method of airfoil based on deep learning was proposed, and the pressure distribution image and airfoil curve were taken as the learning objects to build the prediction model. It can obtain the airfoil curve by using convolutional neural network to extract geometric features from pressure distribution image, while avoiding the time-consuming process of traditional numerical calculation method. In the test case, 6 000 samples of pressure distribution images and airfoil curves were used to train the prediction model, and the other 561 samples were used for validation with validation time only 67 s. The average relative error between the predicted and CFD results was 055%. In the comparison test, noise was added to the pressure distribution curve and the output layer size was changed to further validate the performance of the model. The results show that the proposed inverse design method of airfoil has strong robustness and prediction accuracy. The method can greatly reduce the computation time and improve the design efficiency of airfoil with satisfactory design accuracy.
- inverse design method /
- airfoil curve /
- pressure distribution /
- ,convolutional neural network /
- deep learning /
- prediction model

HTML

参考文献(39)

[1]	王清,招启军.基于遗传算法的旋翼翼型综合气动优化设计［J］.航空动力学报,2016,31(6):1486-1495. WANG Qing,ZHAO Qijun.Synthetical optimization design of rotor airfoil by genetic algorithm［J］.Journal of Aerospace Power,2016,31(6):1486-1495.(in Chinese)
[2]	高正红,王超.飞行器气动外形设计方法研究与进展［J］.空气动力学学报,2017,35(4):516-528. GAO Zhenghong,WANG Chao.Aerodynamic shape design methods for aircraft:status and trends［J］.Acta Aerodynamica Sinica,2017,35(4):516-528.(in Chinese)
[3]	白俊强,邱亚松,华俊.改进型Gappy POD翼型反设计方法［J］.航空学报,2013,34(4):762-771. BAI J Qunqiang,QIU Yasong,HUA Jun.Improved airfoil inverse design method based on Gappy POD［J］.Acta Aeronautica et Astronautica Sinica,2013,34(4):762-771.(in Chinese)
[4]	赵国庆,招启军.基于目标压力分布的旋翼先进气动外形反设计分析方法［J］,航空学报,2014,35(3):744-755. ZHAO Guoqing,ZHAO Qijun.Inverse design analysis method on rotor with advanced aerodynamic configuration based upon target pressurse distribution［J］.Acta Aeronautica et Astronautica Sinica,2014,35(3):744-755.(in Chinese)
[5]	LIGHTHILL M J.A new method of two-dimensional aerodynamics design［R］.Aeronautical Research Council,ARC Technical Report No.2112,1945.
[6]	TAKANASHI S.Iterative three-dimensional transonic wing design using integral equations［J］.Journal of Aircraft,1985,22(8):655-660.
[7]	詹浩,华俊,张仲寅.基于余量修正原理的多翼面气动力反设计方法［J］.航空学报,2003,24(5):411-413. ZHAN Hao,HUA Jun,ZHANG Zhongyin.Design of multi-lifting surfaces based on iterative residual correction［J］.Acta Aeronautica et Astronautica Sinica,2003,24(5):411-413.(in Chinese)
[8]	华俊,张仲寅．一种跨声速翼型设计方法及设计诸例［J］．空气动力学学报,1990,8(2):117-123. HUA Jun,ZHANG Zhongyin.A Transonic airfoil design method and example［J］.Acta Aerodynamica Sinica,1990,8(2):117-123.(in Chinese)
[9]	李焦赞.基于目标压力分布优化翼型反设计方法研究［D］.西安:西北工业大学,2007. LI Jiaozan.Study on inverse design method of airfoil based on optimization of target pressure distribution［D］.Xi’an:Northwestern Polytechnical University,2007.(in Chinese)
[10]	赵小虎,阎超.基于气动数值模拟的翼型反设计方法［J］.航空学报,1997,18(6):9-12. ZHAO Xiaohu,YAN Chao.Aerodynamic inverse design method of airfoil via CFD［J］.Acta Aeronautica et Astronautica Sinica,1997,18(6):9-12.(in Chinese)
[11]	BUI-THANH T,DAMODARAN M,WILLCOX K.Aerodynamic data reconstruction and inverse design using proper orthogonal decomposition［J］.AIAA Journal,2004,42(8):1501-1516.
[12]	刘俊,宋文萍,韩忠华.Kriging模型在翼型反设计中的应用研究［J］.空气动力学学报,2014,32(4):518-526. LIU Jun,SONG Wenping,HAN Zhonghua,et al.Kriging-based airfoil inverse design［J］.Acta Aerodynamica Sinica,2014,32(4):518-526.(in Chinese)
[13]	韩少强,宋文萍,韩忠华,等.基于梯度增强型Kriging模型的气动反设计方法［J］.航空学报,2017,38(7):138-152. HAN Shaoqiang,SONG Wenping,HAN Zhonghua,et al.Aerodynamic inverse design method based on gradient-enhanced Kriging model［J］.Acta Aeronautica et Astronautica Sinica,2017,38(7):138-152.(in Chinese)
[14]	詹浩,朱军,白俊强,等.遗传算法结合反设计的翼型优化设计研究［J］.西北工业大学学报,2006,24(5):541-543. ZHAN Hao,ZHU Jun,BAI Junqiang,et al.Improving design of airfoil with genetic algorithm (GA) combined with inverse design method［J］.Journal of Northwestern Polytechnical University,2006,24(5):541-543.(in Chinese)
[15]	李秀娟,廖文和,刘浩.基于翼型反设计的遗传算法［J］.南京航空航天大学学报,2007,39(2):263-266. LI Xiujuan,LIAO Wenhe,LIU Hao.Genetic algorithms based on airfoil inverse design［J］.Journal of Nanjing University of Aeronautics and Astronautics,2007,39(2):263-266.(in Chinese)
[16]	WANG X,DAMODARAN M,LEE S L.Inverse transonic airfoil design using parallel simulated annealing and computational fluid dynamics［J］.AIAA Journal,2002,40(4):791-794.
[17]	单志辉.基于高斯过程回归的翼型快速设计研究［D］.南京:南京航空航天大学,2011. SHAN Zhihui.Fast airfoil design based on gaussian process regression［D］.Nanjing:Nanjing University of Aeronautics and Astronautics,2011.(in Chinese)
[18]	刘凌君,周越,高振勋.基于神经网络的翼型气动力计算和反设计方法［J］.气体物理,2018,3(5):43-49. LIU Lingjun,ZHOU Yue,GAO Zhenxun.Aerodynamic force calculation and inverse design for airfoil based on neural network［J］.Physics of Gases,2018,3(5):41-47.(in Chinese)
[19]	SUN G,SUN Y,WANG S.Artificial neural network based inverse design:airfoils and wings［J］.Aerospace Science and Technology,2015,42:415-428.
[20]	SOBIECZKY H.Parametric airfoils and wings［C］∥Proceedings of Recent Development of Aerodynamic Design Methodologies.Berlin:Vieweg Teubner Verlag,1999:71-87.
[21]	张德虎,席胜,田鼎.典型翼型参数化方法的翼型外形控制能力评估［J］.航空工程进展,2014,5(3):281-288. ZHANG Dehu,XI Sheng,TIAN Ding.Geometry control ability evaluation of classical airfoil parametric methods［J］.Advance in Aeronautical Science And Engineering,2014,5(3):281-288.(in Chinese)
[22]	SAMAREH J A,SAMAREH J A,POLYNOMIAL B B.A survey of shape parameterization techniques［R］.NASA/CP-1999-209136,1999.
[23]	HINTON G E,SALAKHUTDINOV R R.Reducing the dimensionality of data with neural networks［J］.Science,2006,313(5786):504-507.
[24]	吴正文.卷积神经网络在图像分类中的应用研究［D］.成都:电子科技大学,2015. WU Zhengwen.Application research of convolution neural network in image classification［D］.Chengdu:University of Electronic Science and Technology,2015.(in Chinese)
[25]	袁江林,郭志刚,陈刚,等.基于深度学习的文本自动生成技术研究综述［J］.信息工程大学学报,2018,19(5):616-620. YUAN Jianglin,GUO Zhigang,CHEN Gang,et al.Summary of text auto-generation technology research based on deep learning［J］.Journal of Information Engineering University,2018,19(5):616-620.(in Chinese)
[26]	FRANK S,GANG L,Dong Y.Conversational speech transcription using context-dependent deep neural networks［R］.Florence,Italy:the 12th Annual Conference of the International Speech Communication Association,2011.
[27]	陈海,钱炜祺,何磊.基于深度学习的翼型气动系数预测［J］.空气动力学学报,2018,36(2):294-299. CHEN Hai,QIAN Weiqi,HE Lei.Aerodynamic coefficient prediction of airfoils based on deep learning［J］.Acta Aerodynamica Sinica,2018,36(2):294-299.(in Chinese)
[28]	廖鹏,姚磊江,白国栋,等.基于深度学习的混合翼型前缘压力分布预测［J］.航空动力学报,2019,34(8):1751-1758. LIAO Peng,YAO Leijiang,BAI Guodong,et al.Prediction of hybrid airfoil leading edge pressure distribution based on deep learning［J］.Journal of Aerospace Power,2019,34(8):1751-1758.(in Chinese)
[29]	LEE S,YOU D.Data-driven prediction of unsteady flow over a circular cylinder using deep learning［J］.Journal of Fluid Mechanics,2019,879:217-254.
[30]	李彦冬,郝宗波,雷航.卷积神经网络研究综述［J］.计算机应用,2016,36(9):2508-2515. LI Yandong,HAO Zongbo,LEI Hang.Survey of convolutional neural network［J］.Journal of Computer Applications,2016,36(9):2508-2515.(in Chinese)
[31]	LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition［J］.Proceedings of the IEEE,1998,86(11):2278-2324.
[32]	LECUN Y,BENGIO Y,HINTON G.Deep learning［J］.Nature,2015,521(7553):436-446.
[33]	GLOROT X,BENGIO Y.Understanding the difficulty of training deep feedforward neural networks［R］.Brookline,MA,USA:International Conference on Artificial Intelligence and Statistics,2010.
[34]	NAIR V,HINTON G E.Rectified linear units improve restricted Boltzmann machines［C］∥Proceedings of the 27th International Conference on Machine Learning.Haifa,Israel:Omni Press,2010:807-814.
[35]	KRiZHEVSKY A,SUTSKEVER I,HINTON G E.Imagenet classificationwith deep convolutional neural networks［C］∥Advances in Neural Information Processing Systems.New York:Curran Associates,2012:1097-1105.
[36]	王建军,高正红.HicksHenne翼型参数化方法分析及改进［J］.航空计算技术,2010,40(4):46-49. WANG Jianjun,GAO Zhenghong.Analysis and improvement of hickshenne airfoil parameterization method［J］.Aeronautical Computing Technique,2010,40(4):46-49.(in Chinese)
[37]	陈作斌,江雄,周铸,等.计算流体技术及应用［J］.中国科学:E辑,2008,38(5):657-670. CHEN Zuobin,JIANG Xiong,ZHOU Zhu,et al.Computational fluid technology and application［J］.Science in China:E,2008,38(5):657-670.(in Chinese)
[38]	王建涛,易贤,肖中云,等.ARJ21-700飞机冰脱落数值模拟［J］.空气动力学学报,2013,31(4):430-436. WANG Jiantao,YI Xian,XIAO Zhongyun,et al.Numerical simulation of ice shedding from ARJ21-700［J］.Acta Aerodynamica Sinica,2013,31(4):430-436.(in Chinese)
[39]	KINGMA D P,BA J.Adam:a method for stochastic optimization［R］.San Diego,USA:the 3rd International Conference for Learning Representations,2015.