Repozytorium publikacji - Politechnika Gdańska

Ever increasing performance requirements make the design of contemporary antenna systems a complex and multi-stage process. One of the challenges, pertinent to the emerging application areas but also some of the recent trends (miniaturization, demands for multi-functionality, etc.), is the necessity of handling several performance figures such as impedance matching, gain, or axial ratio, often over multiple frequency bands. The fundamental difficulty is that most of the design objectives are at least partially conflicting. Hence, an improvement of one generally implies degradation of the others. The knowledge of available trade-offs is indispensable and can be acquired through multi-objective optimization (MO). Unfortunately, MO is computationally expensive when executed at the level of EM simulation models, otherwise necessary from the standpoint of antenna evaluation reliability. This paper proposes a computationally efficient framework for MO of antennas. Its keystone is the recently introduced nested kriging modeling technology, here adopted for identifying the design space region that contains the best design trade-offs, as well as for constructing a fast surrogate model to be processed by the MO algorithm. The technique is demonstrated through a two-objective optimization of a planar Yagi antenna (with respect to the impedance matching and gain enhancement) and three-objective design of a compact wideband antenna (with respect to the impedance matching, gain variability, and the footprint area). In both cases, the Pareto set is obtained at the low cost of a few hundred of antenna simulations, even though the optimization process is exclusively based on high-fidelity EM analysis.

Autorzy

• prof. dr inż. Sławomir Kozieł link otwiera się w nowej karcie ,
• dr hab. inż. Anna Pietrenko-Dąbrowska link otwiera się w nowej karcie