The aerodynamic design of modern civil aircraft requires a true sense of intelligence since it requires a good understanding of transonic aerodynamics and sufficient experience.Reinforcement learning is an artificial general intelligence that can learn sophisticated skills by trial-and-error, rather than simply extracting features or making predictions from data.The present paper utilizes a deep reinforcement learning algorithm to learn the policy for reducing the aerodynamic drag of supercritical airfoils. The policy is designed to take actions based on features of the wall Mach number distribution so that the learned policy can be more general. The initial policy for reinforcement learning is pretrained through imitation learning, and the result is compared with randomly generated initial policies. The policy is then trained in environments based on surrogate models, of which the mean drag reduction of 200 airfoils can be effectively improved by reinforcement learning. The policy is also tested by multiple airfoils in different flow conditions using computational fluid dynamics calculations. The results show that the policy is effective in both the training condition and other similar conditions, and the policy can be applied repeatedly to achieve greater drag reduction.

No abstract

Effective access to obtain the complex flow fields around an airfoil is crucial in improving the quality of supercritical wings. In this study, a systematic method based on generative deep learning is developed to extract features for depicting the flow fields and predict the steady flow fields around supercritical airfoils. To begin with, a variational autoencoder (VAE) network is designed to extract representative features of the flow fields. Specifically, the principal component analysis technique is adopted to realize feature reduction, aiming to obtain the optimal dimension of features in VAE. Afterward, the extracted features are incorporated into the dataset, followed by the mapping from the airfoil shapes to features via a multilayer perception (MLP) model. Eventually, a composite network is adopted to connect the MLP and the decoder of VAE for predicting the flow fields given the airfoil. The proposed VAE network achieves compression of high-dimensional flow field data into ten representative features. The statistical results indicate the accurate and generalized performance of the proposed method in reconstructing and predicting flow fields around a supercritical airfoil. Especially, our method obtains accurate prediction results over the shock area, indicating its superiority in conducting turbulent flow under high Reynolds number.

No abstract

No abstract