![]() 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. The airfoils are listed alphabetically by the airfoil filename (which is usually close to the airfoil name). The UIUC Airfoil Data Site gives some background on the database. The policy is also tested by multiple airfoils in different flow conditions using computational fluid dynamics calculations. Included below are coordinates for approximately 1,600 airfoils (Version 2.0). 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 initial policy for reinforcement learning is pretrained through imitation learning, and the result is compared with randomly generated initial policies. 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 parameters in CST are perturbed in the region of 0.05, 0. In addition, referring to the aerodynamic shape optimization settings in 15, 28, a CST method with l 7 is used to deform supercritical airfoil shapes. Correlation screening and multivariate regression are carried out to discover knowledge about the airfoil drag divergence Mach number and pressure. The airfoils are in the the updates directory. In this work, typical supercritical airfoil RAE2822 is used as baseline airfoil. This paper generates a supercritical airfoil database that covers the typical free stream Mach number, angle of attack, lift coefficient, and geometry of modern transonic commercial aircraft. Coordinates for these airfoils were first included in LSATS Vol 5 but somehow they slipped by there inclusion here in the airfoil coordinates database. The present paper utilizes a deep reinforcement learning algorithm to learn the policy for reducing the aerodynamic drag of supercritical airfoils. This update includes the Chris Lyon CAL2263m, CAL4014l, and CAL1215j airfoils. 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. Download a PDF of the paper titled Learning the aerodynamic design of supercritical airfoils through deep reinforcement learning, by Runze Li and 2 other authors Download PDF Abstract: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.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |