An Evolutionary Architecture for the Automated Conceptual Design of Aerospace Systems
Michael Duffy, Soon‐Jo Chung, Lawrence Bergman · Infotech@Aerospace 2011 · 2011
Many different approaches to total system design and optimization have been atempted. The Genetic Learning Automated Design Optimization Software (GLADOS) presented here represent a flexible evolutionary algorithm bases architecture intented to allow for the generation of conceptual or preliminary design stage aircraft designs without any human beings in the loop. The benefits of this approach being more thorough exploration of the design space, the ability to analyze and produce high commonality and modular and modular design configurations, and allowing design capabilities to directly growing computational capabilities. Both an overview of the idealized architecture as well as the results from a simplified test version are presented.