OATAO - Open Archive Toulouse Archive Ouverte Open Access Week

Active control of digital morphing airfoilsusing deep learning

Morlier, Joseph and Carreira Rufato, Raul Active control of digital morphing airfoilsusing deep learning. ( In Press: 2021) In: Aerobest 2021, 21 July 2021 - 23 July 2021 (Online conference, Portugal).

[img] (Document in English)

PDF - Depositor and staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
633kB

Abstract

Detect and prevent an aircraft instability condition is extremely important, especially for flight control, and morphing airfoils can be used for this purpose. This work proposes the determination of a digital morphing airfoil, using a deep learning approach, to avoid an unstable aeroelastic condition in a 2D wing model. To parametrize the airfoil’s geometry, Bezier – Parsec 3434 parametrization was used and some of the parameters were determined by an optimization process based on a Genetic Algorithm. The airfoil’s geometric Cg position, cl, cd and cm distributions for some angles of attack were used to train a deep neural network, capable to estimate the desired BP3434 parameters. Finally, this trained machine learning model was then coupled to the 2D aeroelastic model of a wing to change the airfoil’s curvature when it faced an instability. The trained deep learning algorithm had an excellent Person’s coefficient of 0.919 when predicting a new geometry. Our methodology permits to automatically detect and avoid instability using digital morphing techniques coupled with AI, using only one sensor, monitoring the dynamic behavior of the airfoil.

Item Type:Conference or Workshop Item (Paper)
Audience (conference):International conference proceedings
Uncontrolled Keywords:
Institution:Université de Toulouse > Institut Supérieur de l'Aéronautique et de l'Espace - ISAE-SUPAERO (FRANCE)
Laboratory name:
Statistics:download
Deposited On:10 Sep 2021 07:50

Repository Staff Only: item control page