28th Congress of the International Council of the Aeronautical Sciences, 23 - 28 September 2012, Brisbane, Australia
Paper ICAS2012-11.10ST


COMPARISON OF INFILL SAMPLING CRITERIA IN KRIGING-BASED AERODYNAMIC OPTIMIZATION

J. Liu, Z. H. Han, W. P. Song
NWPU, Xi'an, China

Keywords: kriging, infill sampling criteria (isc), optimization

This paper compared 6 infill sampling criteria in Kriging-based aerodynamic optimization: minimizing the predictor (MP), lower confidence bounding, expected improvement (EI), probability of improvement, maximizing the predicted error, and combining EI with MP. The results show that EI performs slightly better than all the others, and ME performs worst.


view full paper