Deep Reinforcement Learning for Flow Control Exploits Different Physics for Increasing Reynolds Number Regimes

• Published in: Actuators Vol. 11; no. 12; p. 359
• Main Authors: Varela, Pau, Suárez, Pol, Alcántara-Ávila, Francisco, Miró, Arnau, Rabault, Jean, Font, Bernat, García-Cuevas, Luis Miguel, Lehmkuhl, Oriol, Vinuesa, Ricardo
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.12.2022
• Subjects: Active control; Actuation; Aeronautics; Aircraft; Artificial neural networks; Boundary layer; Boundary layers; Control algorithms; Control methods; Cylinders; Deep learning; deep reinforcement learning; Drag reduction; Emissions; Flow control; Flow separation; Flow simulation; Fluid flow; Machine learning; Neural networks; numerical simulation; Optimization; Partial differential equations; Reynolds number; Two dimensional flow; Velocity; wake dynamics

The increase in emissions associated with aviation requires deeper research into novel sensing and flow-control strategies to obtain improved aerodynamic performances. In this context, data-driven methods are suitable for exploring new approaches to control the flow and develop more efficient strate...