查看文章

THE optimization of aerospace systems is an iter-ative process that requires computational models embodied in complex and expensive analysis software. This paradigm is well exemplified by the field of Multidisciplinary Design Optimization (MDO) which attempts to exploit the synergism of mutually interacting disciplines in order to improve the performance of a given design, while increasing the level of confidence that the designer places on the outcome of the design itself. MDO methods greatly increase the computational burden and complexity of the design process. 1'2 For this reason, high-fidelity analysis software typically used in single discipline designs may not be suitable for direct use in MDO. 3 Faced with these problems, the alternative of using approximation models of the actual analysis software has received increased attention in recent years. A second advantage of using approximation models during …