Description du projet
Des stratégies innovantes de contrôle du débit pour réduire les émissions
Le secteur du transport aérien a des répercussions économiques et environnementales importantes en raison de ses fortes émissions de CO2. Le besoin d’améliorer les performances aérodynamiques des ailes d’avion est urgent pour réduire la consommation de carburant et les émissions. L’une des stratégies consiste à effectuer un contrôle de flux. Le projet DEEPCONTROL, financé par le CER, utilisera des simulations haute fidélité et l’apprentissage par renforcement profond pour développer un cadre permettant de prédire et de contrôler en temps réel l’écoulement autour des sections d’ailes et des ailes tridimensionnelles, en se basant uniquement sur des mesures éparses. Le projet veut découvrir de nouvelles solutions en termes d’actionnement du flux et de conception de la géométrie des ailettes pour améliorer la durabilité du secteur. DEEPCONTROL réalisera également des expériences détaillées en soufflerie à l’Institut royal de technologie KTH, en Suède, afin d’évaluer le cadre pour les applications en temps réel.
Objectif
Over the past decades, aviation has become an essential component of today’s globalized world: before the current pandemic of coronavirus disease 2019 (COVID-19), over 100,000 flights took off everyday worldwide, and a number of studies indicate that after the pandemic its relevance in the transportation mix will be similar to that before COVID-19. Aviation alone is responsible for 12% of the carbon dioxide emissions from the whole transportation sector, and for 3% of the total CO2 emissions in the world. Due to the major environmental and economical impacts associated to aviation, there is a pressing need for improving the aerodynamic performance of airplane wings to reduce fuel consumption and emissions. This implies reducing the force parallel to the incoming flow, i.e. the drag, and one of the strategies to achieve such a reduction is to perform flow control.
DEEPCONTROL aims at using high-fidelity simulations and deep reinforcement learning to develop a framework for real-time prediction and control of the flow around wing sections and three-dimensional wings based only on sparse measurements. We will first perform high-order spectral-element simulations of wing sections and three-dimensional wings at high Reynolds numbers. Using sparse measurements at the wall, we will reconstruct the velocity fluctuations above the wall within a region of interest. To this end, we will employ a generative adversarial network (GAN), together with a fully-convolutional network (FCN) and modal decomposition. Then, we will perform flow control based on deep reinforcement learning (DRL), which will enable discovering novel solutions in terms of flow actuation and design of winglet geometry. In order to assess the robustness of the framework for real-time applications, we will carry out detailed wind-tunnel experiments at KTH.
This framework will constitute a breakthrough in aviation sustainability, and will enable developing more efficient aeronautical solutions worldwide.
Champ scientifique
- medical and health scienceshealth sciencespublic healthepidemiologypandemics
- natural sciencescomputer and information sciencesartificial intelligencemachine learningreinforcement learning
- natural sciencesphysical sciencesclassical mechanicsfluid mechanicsfluid dynamics
- medical and health scienceshealth sciencesinfectious diseasesRNA virusescoronaviruses
- engineering and technologyenvironmental engineeringenergy and fuels
Programme(s)
- HORIZON.1.1 - European Research Council (ERC) Main Programme
Régime de financement
HORIZON-ERC - HORIZON ERC GrantsInstitution d’accueil
100 44 Stockholm
Suède