Multi-objective design of compact RF/microwave components using decomposition and surrogate modeling

A technique for computationally efficient multi-objective design optimization of compact RF/microwave components is presented. Our approach exploits response surface approximation (RSA) models of the basic building blocks of the structure of interest and their cascade connection as a fast low-fidelity model of the entire structure. The set of Pareto-optimal solutions (corresponding to the best possible trade-offs between conflicting objectives) is obtained by optimizing the low-fidelity model using a multi-objective evolutionary algorithm. The Pareto set is then iteratively refined through re-optimization of the low-fidelity model corrected by means of space mapping, and using EM-simulation data of the structure at hand sampled along the current approximation of the Pareto front. The proposed methodology is demonstrated using two examples of compact wideband matching transformers.