By Slawomir Koziel, Stanislav Ogurtsov

This short stories a few recommendations exploiting the surrogate-based optimization suggestion and variable-fidelity EM simulations for effective optimization of antenna buildings. The advent of every process is illustrated with examples of antenna layout. The authors display the ways that practitioners can receive an optimized antenna layout on the computational expense reminiscent of a couple of high-fidelity EM simulations of the antenna constitution. there's additionally a dialogue of the choice of antenna version constancy and its impact on functionality of the surrogate-based layout approach. This quantity is acceptable for electric engineers in academia in addition to undefined, antenna designers and engineers facing computationally-expensive layout difficulties.

**Read or Download Antenna Design by Simulation-Driven Optimization PDF**

**Best mathematics books**

- Numerical solution of elliptic equations (SIAM, 1972)(ISBN 0686242513)
- Kaltgasspritzen von Kupfer. Einestromungsmechanische und werkstoffkundliche Analyse und Optimierung des Spritzprozesses
- H-n-perturbations of Self-adjoint Operators and Kreins Resolvent Formula
- Partial Differential Equations in Classical Mathematical Physics (CUP 1998)
- War Stories From Applied Math
- Professor Stewart's Cabinet of Mathematical Curiosities

**Extra info for Antenna Design by Simulation-Driven Optimization**

**Example text**

The main advantage of physics-based models is that—because of exploiting some knowledge embedded in the low-fidelity model—a limited amount of high- fidelity data is necessary to ensure decent accuracy. , they can provide reliable prediction of the high-fidelity model response at the designs not used in the training process. These advantages are normally translated into better efficiency (in particular, lower CPU cost) when physics-based surrogates are used in the design optimization process (Koziel et al.

12) where βi(x) = βi(x(i)) + ∇β(x(i))T(x − x(i)) and where β(x) = f(x)/c(x). , agreement of function values and their gradients at x(i) (Alexandrov and Lewis 2001). Another way of correcting the low-fidelity model is so-called input space mapping (ISM) (Bandler et al. 13) with the model parameters c(i) obtained by minimizing ||Rf(x(i)) − Rc(x(i) + c(i))||. 5 shows an example of a filter structure evaluated using EM simulation (high-fidelity model), its circuit equivalent (low-fidelity model), and the corresponding |S21| responses before and after applying the ISM correction.

2004a, b). In case of antennas, the only universally available way of obtaining low-fidelity models is through coarse-discretization EM simulation. A discussion of low-fidelity antenna models is presented in Chap. 5 of this book. The main advantage of physics-based models is that—because of exploiting some knowledge embedded in the low-fidelity model—a limited amount of high- fidelity data is necessary to ensure decent accuracy. , they can provide reliable prediction of the high-fidelity model response at the designs not used in the training process.