Artificial neural network based wing planform aerodynamic optimization

Abstract

Purpose Environmental and operational restrictions increasingly drive modern aircraft design due to the growing impact of global warming on the ecology. Regulations and industrial measures are being introduced to make air traffic greener, including restrictions and environmental targets for aircraft design that increase aerodynamic efficiency. This study aims to maximize aerodynamic efficiency by identifying optimal values for sweep angle, taper ratio, twist angle and wing incidence angle parameters in wing design while keeping wing area and span constant. Design/methodology/approach Finding optimal wing values by using gradient-based and evolutionary algorithm methods is very time-consuming. Therefore, an artificial neural network-based surrogate model was developed. Computational fluid dynamics (CFD) analyses were carried out by using Reynolds-averaged Navier–Stokes equations to create a properly trained data set using a feedforward neural network. Findings The results showed how a wing could be optimized by using a CFD-based surrogate model. The two optimum results obtained resulted in increases of 10.7397% and 10.65% in the aerodynamic efficiency of the baseline design ONERA M6 wing. Originality/value The originality of this study lies in the combination of sweep angle, taper ratio, twist angle and wing incidence angle within the scope of wing optimization calculations.