A Dynamic Sampling Method for Kriging and Cokriging Surrogate Models
Markus P. Rumpfkeil, Wataru Yamazaki, Mavriplis Dimitri · 49th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition · 2011
In this paper we describe our gradient and Hessian enhanced Kriging surrogate model with dynamic sample point selection. We demonstrate the quality of the surrogate by comparison with higher-dimensional analytic test functions. We also apply the surrogate model to uncertainty quantification and robust optimization problems using inexpensive Monte-Carlo simulations. All applications benefit from the additional gradient and Hessian information as well as the dynamic sample point selection by requiring fewer function evaluations and overall less computational time.