Suscripción a Biblioteca: Guest
Portal Digitalde Biblioteca Digital eLibros Revistas Referencias y Libros de Ponencias Colecciones
TsAGI Science Journal

ISSN Imprimir: 1948-2590
ISSN En Línea: 1948-2604

TsAGI Science Journal

DOI: 10.1615/TsAGISciJ.2012004704
pages 817-828

APPLICATION OF NEURAL NETWORKS IN THE SIMULATION OF DYNAMIC EFFECTS OF CANARD AIRCRAFT AERODYNAMICS

Dmitry Igorevich Ignatyev
Central Aerohydrodynamic Institute (TsAGI), Zhukovsky str. 1, Zhukovsky, Moscow region, 140180 Russia
Alexander Nikolaevich Khrabrov
Central Aerohydrodynamic Institute (TsAGI), Zhukovsky str. 1, Zhukovsky, 140180, Russia

SINOPSIS

The possibility of the application of neural networks in the simulation of the pitch moment coefficient in a wide range of angles of attack for harmonic oscillations of transonic aircraft is shown. TsAGI carried out an experimental study of transonic aircraft in the project SimSAC of the European 6th Framework Program. Traditional simulation of the unsteady aerodynamic characteristics for this aircraft fails to yield the required precision at high angles of attack due to the significant dynamic effects caused by canard flow separation dynamics. Neural network simulation of the pitch moment versus angle of attack in the form of hysteresis loops was carried out in a wide range of angles of attack. The dynamic derivatives of the pitch moment were simulated in a wide range of angles of attack. The frequency of the model oscillation was taken into account.