Accelerated Multi-Objective Design Optimization of Antennas By Surrogate Modeling and Domain Segmentation

Multi-objective optimization yields indispensable information about the best possible design trade-offs of an antenna structure, yet it is challenging if full-wave electromagnetic (EM) analysis is utilized for performance evaluation. The latter is a necessity for majority of contemporary antennas as it is the only way of achieving acceptable modeling accuracy. In this paper, a procedure for accelerated multi-objective design of antennas is proposed that exploits fast data-driven surrogates constructed at the level of coarse-discretization EM simulations, multi-objective evolutionary algorithm to yield an initial approximation of the Pareto set, and response correction methods for design refinement (i.e., elevating the selected Pareto-optimal designs to the high-fidelity EM simulation model level). To reduce the computational cost of setting up the surrogate, the relevant part of the design space (i.e., the part containing the Pareto front) is first identified through a series of single-objective optimization runs and subsequently represented by a set of adjacent compartments with separate surrogate models established within them. This segmentation process dramatically reduces the number of training samples necessary to build an accurate model thus limiting the overall optimization cost. Our approach is demonstrated using a UWB monopole antenna and compared to a state-of-the-art surrogate-assisted technique that does not use domain segmentation.