Optimization of high-speed electromagnetic systems with accurate parametric macromodels generated using sequential sampling of the design space

This paper presents a design optimization approach for electromagnetic systems using parametric macromodels. The parametric macro-models are generated using an efficient sequential sampling of the design space of interest which ensures optimal sample selection for a required level of accuracy. The proposed method is validated on a microwave notch filter example for which the parametric macromodel is used in a minimax optimization algorithm so that the design parameters are optimized for some specific electrical design performances.