An efficient computational framework is presented and applied to the inverse aerodynamic shape design problem. The main building block is a novel neural network capable to accurately predict the pressure distribution on aerofoils and wings. The trained neural network is used to accelerate the evaluation of the objective function in an optimisation algorithm based on the gradient-free modified cuckoo search method. Two applications are presented in two and three dimensions for problems involving up to 50 geometric parameters.
Inverse Aerodynamic Design Using Neural Networks
Computational Methods
Advances in Computational Methods and Technologies in Aeronautics and Industry ; Chapter : 10 ; 131-143
2022-12-13
13 pages
Article/Chapter (Book)
Electronic Resource
English
Aerodynamic Design Using Neural Networks
AIAA | 2000
|Aerodynamic Design Using Neural Networks
British Library Conference Proceedings | 1998
|Aerodynamic design using neural networks
AIAA | 1998
|Information Sciences - Aerodynamic Design Using Neural Networks
Online Contents | 2003