Department of Design and Production of Radio Electronic Equipment

National Technical University of Ukraine
"Igor Sikorsky Kyiv Polytechnic Institute"

Electromagnetic crystal filters modeling using artificial neural networks

Author: Kovbych Anton

Group: РІ-62м

Supervisor: Adamenko Yulia

ELK: Master

Speciality: 172 Telecommunications and Radio Engineering

Specialization: Intelligent Technologies of Microsystem Radioelectronic Equipment

Summary: The object of the study is electromagnetic crystal heterogeneity.
The subject of research is the synthesis of electromagnetic crystals by artificial neural network.
The purpose of the work is to study the possibility of using artificial neural networks to create the structure of electromagnetic crystal heterogeneities and to confirm this possibility in practice.
The theoretical foundations of electromagnetic crystals, artificial neural networks, possibilities and peculiarities of their construction on a computer are considered, existing solutions for simplification and automation of this process are considered.
As a result of the work, models of the artificial neural network for individual heterogeneities have been developed, which allow obtaining geometric sizes of structures according to the desired frequency characteristics. The work of these models is checked.
The results of the master's thesis will become the basis for further research in the direction of synthesis of structures of frequency-selective devices based on electromagnetic crystals according to the given characteristics with the help of artificial neural networks.
On the theme of the master's dissertation the following works were published:
Abstracts in the framework of the seventh international scientific and technical conference "Radio engineering fields, signals, devices and systems" in 2018 entitled "Synthesis of quasistricated reactive elements based on crystal-like structures using artificial neural network".
Abstracts in the framework of the first All-Ukrainian Scientific and Technical Conference of Students and Postgraduates "Radio Electronics in the XXI Century" in 2018 entitled "Synthesis of Inductive Electromagnetic Crystal Inhomogeneity Using the Artificial Neural Network"

Keywords: electromagnetic crystal, electromagnetic crystal heterogeneity, artificial neural network, bandpass filter, microwave, microstrip line, cst microwave studio

The work consists of:  101 pages of explanatory note and 2 appendix

Explanatory note: Full text