Abstract Multidisciplinary design optimizations are practical key technology for the efficient design of a supersonic transport (SST). However, the computational cost should be expensive with high fidelity flow solver. Thus, the surrogate models such as the Kriging method is one of the promising technique. On the other hand, the computational cost still expensive, because a lot of CFD runs is required to achieve global search with high-fidelity solver. In this study, to develop higher efficient global exploration method, it was considered that a fusion of the database by a low cost/low fidelity solver and a high cost/high fidelity solve using Kriging model. A test problem and a design problem of the wing of SST was carried out to investigate the efficiency of the proposed method. In the design of the SST, the liner potential solver and the structured Euler solver was employed. According to the result, the total computational cost was drastically reduced while the same optimum solution can be explored as a single-fidelity optimization.
[1]
C. Fonseca,et al.
GENETIC ALGORITHMS FOR MULTI-OBJECTIVE OPTIMIZATION: FORMULATION, DISCUSSION, AND GENERALIZATION
,
1993
.
[2]
Shigeru Obayashi,et al.
Multi-Objective Design Exploration and Its Application to Regional-Jet Wing Design
,
2007
.
[3]
Peter J. Fleming,et al.
Genetic Algorithms for Multiobjective Optimization: FormulationDiscussion and Generalization
,
1993,
ICGA.
[4]
Art B. Owen,et al.
9 Computer experiments
,
1996,
Design and analysis of experiments.
[5]
P. C. Gehlen,et al.
Computer Experiments
,
1996
.
[6]
Takeshi Takatoya,et al.
Design-Informatics Approach for Intimate Configuration of Silent Supersonic Technology Demonstrator
,
2009
.
[7]
Kazuomi Yamamoto,et al.
Efficient Optimization Design Method Using Kriging Model
,
2005
.
[8]
Shigeru Horinouchi.
Conceptual Design of a Low Boom SSBJ
,
2005
.