Conceptual aircraft design: towards multiobjective, robust and uncertain optimisation

Abstract : Aircraft sizing studies consist in determining the main characteristics of an aircraft starting from a set of requirements. These studies can be summarized as global constrained optimisation problems with typically one thousand parameters and almost as many constraints. The constraints express physical feasibility and the requirements to be satisfied, and the objectives are market driven performances of the aircraft. Moreover, aircraft sizing is typically a multicriteria optimisation problem because of some competing objectives.

The aim of this thesis is to introduce new mathematical methods that can be useful in a future project sizing tool to treat the aircraft configuration optimisation problem. We contributed in improving the optimisation methods that are currently used in the Airbus Future Project Office. By using genetic algorithms, we made the mono-criterion optimisation process more robust. Then, we introduced multicriteria optimisation methods because we had several conflicting criteria to consider. As the calculation times became important, we decided to substitute the aircraft model by a surrogate model. We implemented radial basis functions to approximate the constraint and the objective functions. Finally, we propagated the model uncertainty to assess the robustness of the optimisation results and we proposed a possible outcome of the integration of these different techniques in order to yield the engineers a global and operational perception of the design space.
Document type :
Theses
Complete list of metadatas

Cited literature [151 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00367210
Contributor : Céline Badufle <>
Submitted on : Tuesday, March 10, 2009 - 3:44:49 PM
Last modification on : Friday, January 10, 2020 - 9:08:06 PM
Long-term archiving on: Tuesday, June 8, 2010 - 11:19:47 PM

Identifiers

  • HAL Id : tel-00367210, version 1

Citation

Céline Badufle. Conceptual aircraft design: towards multiobjective, robust and uncertain optimisation. Mathematics [math]. Université Paul Sabatier - Toulouse III, 2007. English. ⟨tel-00367210⟩

Share

Metrics

Record views

751

Files downloads

2097