Aircraft configuration development through surrogate-based robust optimization using a real-coded fuzzy-genetic algorithm

• Main Authors: Banal, Lemuel F., Gan Lim, Laurence A., Ubando, Aristotle T., Fernando, Arvin H., Augusto, Gerardo L.
• Format: text
• Published: Animo Repository 2017
• Subjects: Airplanes—Models; Mechanical Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1537; https://animorepository.dlsu.edu.ph/context/faculty_research/article/2536/type/native/viewcontent
• Institution: De La Salle University

An alternative methodology that views aircraft configuration development from an optimization perspective is proposed. The method hinges on the idea that design requirements can be expressed as objectives and constraints, which in turn can be expressed as functions of design variables that define the aircraft configuration. The resulting model will reflect the inherent complexity of the aircraft and it cannot be expected to be accurate especially at such an early stage of the design process. Considering the nature of the problem and the design variables, a real-coded genetic algorithm is used as the solution tool. Fuzzy logic is used to avoid the unwarranted imposition of crisp criteria on the low-fidelity model. It is also used in the evaluation of fitness of individuals. Moreover, principles of robust design are integrated into the algorithm to mitigate the sensitivity of objectives on unavoidable variations in the design variables without actually eliminating the root causes. Robustness of objectives are accounted for through their respective standard deviations computed using a surrogate as embodied by a quadratic response surface model. Compared to the conventional approach which is sequential, the proposed method is able to synthesize certain design steps and simultaneously determine key design parameters. It is also able to output in a single run not just one but a set of fuzzy-Pareto optimal candidate configurations subject for validation and higher-fidelity analysis in the subsequent phases of the design process. The availability of options increases the success rate, reduces design iterations, and facilitates a faster design process. © 2017 IEEE.