Inverse design has long been an efficient and powerful design tool in the aircraft industry.In this paper,a novel inverse design method for supercritical airfoils is proposed based on genera-tive models in deep learning.A Conditional Variational AutoEncoder (CVAE) and an integrated generative network CVAE-GAN that combines the CVAE with the Wasserstein Generative Adver-sarial Networks (WGAN),are conducted as generative models.They are used to generate target wall Mach distributions for the inverse design that matches specified features,such as locations of suction peak,shock and aft loading.Qualitative and quantitative results show that both adopted generative models can generate diverse and realistic wall Mach number distributions satisfying the given features.The CVAE-GAN model outperforms the CVAE model and achieves better recon-struction accuracies for all the samples in the dataset.Furthermore,a deep neural network for non-linear mapping is adopted to obtain the airfoil shape corresponding to the target wall Mach number distribution.The performances of the designed deep neural network are fully demonstrated and a smoothness measurement is proposed to quantify small oscillations in the airfoil surface,proving the authenticity and accuracy of the generated airfoil shapes.

Jing WANG;Runze LI;Cheng HE;Haixin CHEN;Ran CHENG;Chen ZHAI;Miao ZHANG

Shanghai Aircraft Design and Research Institute,Shanghai 200436,China;School of Aerospace Engineering,Tsinghua University,Beijing 100084,China;Guangdong Provincial Key Laboratory of Brain-Inspired Intelligent Computation,Department of Computer Science and Engineering,Southern University of Science and Technology,Shenzhen 518055,China

中国航空学报（英文版）

[1]Jing WANG;Runze LI;Cheng HE;Haixin CHEN;Ran CHENG;Chen ZHAI;Miao ZHANG-.An inverse design method for supercritical airfoil based on conditional generative models)[J].中国航空学报（英文版）,2022(03):62-74

airfoils,Adver

An,inverse,method,supercritical,conditional,generative,models,Inverse,has,long,been,efficient,powerful,tool,aircraft,industry,this,paper,novel,proposed,deep,learning,Conditional,Variational,AutoEncoder,CVAE,integrated,network,that,combines,Wasserstein,Generative,sarial,Networks,WGAN,are,conducted,They,used,target,wall,Mach,distributions,matches,specified,features,such,locations,suction,peak,shock,loading,Qualitative,quantitative,results,show,both,adopted,can,diverse,realistic,number,satisfying,given,outperforms,achieves,better,recon,struction,accuracies,samples,dataset,Furthermore,neural,linear,mapping,obtain,corresponding,performances,designed,fully,demonstrated,smoothness,measurement,quantify,small,oscillations,surface,proving,authenticity,accuracy,generated,shapes

0.521074