Commercial transport aircraft conceptual design using a genetic algorithm based approach

Abstract

Fixed-wing aircraft design is a complex engineering problem, yet the conceptual phase of design is often limited in the number of design variables examined. Further, to begin the design process, many decisions about an aircraft's configuration are based upon qualitative choices of the designer(s). The use of a genetic algorithm (GA) can assist in aircraft conceptual design by reducing the number of qualitative decisions made during the design process while increasing the number of design variables taken into consideration. The genetic algorithm is a search method based on the patterns of natural selection and reproduction common to biological populations. Since the GA operates as a non-calculus based method, discrete and continuous design variables can be handled with equal ease. This paper describes a hybrid approach with the implementation of a GA as a less-biased, automated approach to conceptual aircraft design and the application of CONMIN, a calculus-based method of feasible directions, to refine the results obtained with the GA. Civilian transport class aircraft are the current focus. The resulting optimization-analysis code is used to generate potential conceptual designs for a specified mission. Results from these design efforts are discussed with insight into the use of GAs for conceptual aircraft design.