Multidisciplinary Optimization Strategies Using Evolutionary Algorithms with Application to Aircraft Design
Artur Fouto, Maria Alexandra Gomes, Afzal Suleman · 50th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference · 2009
This paper describes the development of a preliminary aircraft design application employing the concepts of Multidisciplinary Optimization (MDO) and evolutionary algorithms. Aerodynamics, structural analysis and flight performance are the main disciplines considered for preliminary aircraft design. Aerodynamic analysis is performed using a 3D panel method with boundary layer correction and the structural analysis is performed using a finite element method, both by external software packages handled by the developed application. A particle swarm optimizer was developed to handle the MDO problem with a large number of design variables, an Artificial Neural Network (ANN) was investigated to predict the Pareto Front (in a context of Multiobjective Optimization) and as an accelerator for the whole optimization process. The goal of developing an application that is fully independent from user input during the optimization process and is able to interact with external analysis tools was reached and several simple aircraft design optimization problems are solved, in order to demonstrate the advantages of the MDO concept and the developed optimization framework. I.