Application of artificial neural networks for simulation of delta wing aerodynamic characteristics

Аuthors

Ignatyev D. I.

Central Aerohydrodynamic Institute named after N.E. Zhukovsky (TsAGI), 1, Zhukovsky str., Zhukovsky, Moscow Region, 140180, Russia

e-mail: d.ignatyev@mail.ru

Abstract

The conception of aerodynamic derivatives which is widely used for simulation of the unsteady aerodynamic characteristics does not provide the required precision of the results at high angles of attack due to the dynamical effects caused by the dynamics of flow separation and vortices breaking. A possibility of artificial neural networks application for simulation of the lift and pitch moment coefficients within the wide range of the angles of attack is demonstrated in the present work. A short definition of the artificial neural networks is presented. Given in the paper is the description of methods of the unstable aerodynamic characteristics mathematical simulation. Mathematical models describing the results of dynamical experiments with forced oscillations obtained at different frequencies are presented. Hysteresis of the lift and pitch moment coefficients is simulated on an example of a delta wing.

Keywords:

artificial neural networks, delta wing, unstable aerodynamic characteristics, high angles of attack, hysteresis