Surrogate-based optimization and efficient global optimization in particular, is considered for aerodynamic design and analysis to deal with some of the drawbacks of classical direct optimization methods. Design of Experiment methods, optimization algorithms, various surrogate modeling methodologies, adaptive sample refinement strategies, multiple criteria for terminating the refinement procedure and several other techniques, are developed and integrated into a practical optimization framework. To search the design space globally and efficiently, several adaptive sample refinement strategies are studied and compared. Two test cases, minimizing the drag of a NLF0416 airfoil with ten design variables and optimizing the performance of a laminar profile with 26 design variables at two design points, are performed. The results indicate that the developed optimization methodology in combination with the adaptive sample refinement strategies features a good balance between global exploration and local exploitation. Additionally, the effect of different design points on the objective function can be automatically considered in the refinement procedure.
[1]
Thomas J. Santner,et al.
The Design and Analysis of Computer Experiments
,
2003,
Springer Series in Statistics.
[2]
T. Gerhold,et al.
Calculation of Complex Three-Dimensional Configurations Employing the DLR-tau-Code
,
1997
.
[3]
Stefan Görtz,et al.
A framework for surrogate-based aerodynamic optimization
,
2011
.
[4]
S. Bates,et al.
Formulation of the Audze--Eglais uniform Latin hypercube design of experiments
,
2003
.
[5]
M. D. McKay,et al.
A comparison of three methods for selecting values of input variables in the analysis of output from a computer code
,
2000
.
[6]
G. Venter,et al.
An algorithm for fast optimal Latin hypercube design of experiments
,
2010
.
[7]
Shigeru Obayashi,et al.
Efficient global optimization (EGO) for multi-objective problem and data mining
,
2005,
2005 IEEE Congress on Evolutionary Computation.