Multi-Objective Design of Antennas Using Variable-Fidelity Simulations and Surrogate Models

Abstract

A computationally-efficient procedure for multi-objective design of antenna structures is presented. Our approach exploits the multi-objective evolutionary algorithm (MOEA) working with a fast antenna surrogate model obtained with kriging interpolation of coarse-discretization simulation data. Response correction techniques are subsequently applied to refine the designs obtained by MOEA. Our methodology allows us to obtain-at a low computational cost-a set of designs corresponding to various trade-offs between the antenna size and the refection coefficient. Two illustration examples are considered: (i) an UWB monocone with two objectives being reduction of the antenna size and minimization of the antenna reflection coefficient in the bandwidth of interest, and (ii) a planar Yagi antenna with the objectives being an increase of the end-fire gain and minimization of the reflection coefficient, both in the bandwidth of interest.Index Terms-Antenna design, computer-aided design (CAD), evolutionary algorithms, multi-objective optimization, surrogate models.

Recent developments in simulation-driven multi-objective design of antennas

“…Consider a monocone structure [24] that operates in the UWB frequency band. The antenna is fed directly through 50-Ohm coaxial line with Teflon filling and outer diameter of 0.635 mm.…”

“…A detailed description of the decomposition procedure is omitted for the sake of brevity. A more detailed explanation is provided in [24]. The surrogate model R s of the Yagi-Uda antenna is optimized using MOEA and solutions with F 1 ≤ −10 dB are utilized in the refinement procedure (cf.…”

“…Recently, two computationally efficient techniques for multi-objective design optimization of antennas have been proposed [24,25]. Both methods rely on fast response surface approximation (RSA) surrogates created from sampled coarse-discretization EM simulation data, as well as refinement procedures intended to obtain representations of the Pareto-optimal sets at the high-fidelity EM antenna model level.…”

“…Both methods rely on fast response surface approximation (RSA) surrogates created from sampled coarse-discretization EM simulation data, as well as refinement procedures intended to obtain representations of the Pareto-optimal sets at the high-fidelity EM antenna model level. The refinement strategy adopted in [24] is point-by-point identification using designs sampled from the initial Pareto set (obtained by optimizing the RSA) model, and suitable response correction techniques. In [25] refinement is realized by blending sparsely sampled high-fidelity EM model data into the RSA surrogate using Co-kriging [26].…”

A Lexicographic Approach for Multi-Objective Optimization in Antenna Array Design

“…However, such "a-posteriori chosen articulation of preferences" approach requires to repeat a large number of multiple single target optimizations with a variation of the optimization objectives. The creation of the Pareto front is generally a large time consuming task [3][4][5][6][7] and requires a computation time that is unacceptable in case of complex or electrically large structures.…”