Aircraft wing design using GA-based multi-level strategies
Andy J. Keane, Nicola Petruzzelli · 8th Symposium on Multidisciplinary Analysis and Optimization · 2000
The design of the wings for a transonic civil transport aircraft is an extremely complex task. It is normally undertaken over an extended time period and at a variety of levels of complexity. Typically, simple empirical models are used at the earliest stages of concept design, followed by ever more complex methods as the design process proceeds towards the final detailed stages. Moreover, this process is increasingly dominated by computational methods for both analysis (such as CFD) and synthesis (such as optimization). In this paper the further development of the Southampton Wing Design Environment is described. This Problem Solving Environment (PSE) combines multiple analysis methods of differing levels of complexity with powerful search and optimization techniques. A particular focus of the research is the use of evolutionary search strategies, such as genetic algorithms, and of data fusion between the results of various analysis codes when applied to concept design problems.