Creating multivariable response surfaces by the two-stage adaptive sampling techniques and its application in the optimization of grounding grids

A fast and efficient adaptive sampling algorithm to obtain the response surface of multivariable objective function is presented. The algorithm is divided into two stages. First, starting with a low-rank interpolation model, the technique systematically increases the rank by optimally choosing new support points in the areas of the highest error until the required accuracy is achieved. Second, the verification between the fitted result and the exact one is carried out. The optimization of grounding grids is performed to show the effectiveness of the proposed algorithm