Surrogate-assisted evolutionary programming for high dimensional constrained black-box optimization
Rommel G. Regis · 2012
This paper presents a novel surrogate-assisted evolutionary programming (EP) method for high dimensional constrained black-box optimization with many black-box inequality constraints. A cubic radial basis function (RBF) surrogate is used and the resulting RBF-assisted EP outperforms a standard EP, an RBF-assisted penalty-based EP, Stochastic Ranking Evolution Strategy and Scatter Search on a 124-D automotive problem with 68 black-box constraints.