Exploration of Neural Networks in the Design of Microwave Structures
Artificial neural networks can be used for modeling microwave structures in order to obtain computationally efficient models of investigated systems. In conjunction with proper optimization techniques, these neural models can be explored for an efficient full-wave design of microwave structures. Moreover, neural networks can serve as a direct design tool of microwave systems. In this paper, an overview of the so-far published applications of neural networks in microwaves is presented and their exploration in the full-wave design is discussed. Described neural modeling and design is illustrated by modeling and design of frequency-selective surfaces.
Keywordsartificial neural networks, linear random search, frequency-selective surfaces, method of moments
Document typePeer reviewed
Document versionFinal PDF
SourceRadioengineering. 1999, vol. 8, č. 1, s. 34-37. ISSN 1210-2512
- 1999/1