In this letter, a design of experiments technique that permits uniform sampling in constrained domains is proposed. The discussed method is applied to generate training data for construction of fast replacement models (surrogates) of antenna input characteristics. The modeling process is design oriented with the surrogate domain spanned by a set of reference designs optimized with respect to the performance figures and/or operating conditions that are of interest. The reference designs are triangulated, and the resulting simplexes are extended in orthogonal directions. Our methodology is demonstrated using two examples: a dual-band dipole and an ultrawideband monopole. The results indicate that the proposed sampling technique leads to considerable improvement of the surrogate model predictive power as compared to random sampling. Numerical results are supported by application studies (antenna optimization) and experimental validation.
© 2017 IEEE.