Prediction of resonance frequencies of rectangular patch antenna using a multilayer perceptron network

Abstract

In this paper, a novel approach to accurately calculate the resonant frequencies of rectangular microstrip antennas using artificial neural networks (ANN) and the method of moments (MOM) is proposed. The ANN is developed to calculate the real part and the imaginary part of the complex resonant frequency of the antenna. The ANN is designed using multilayer perceptron network (MLP). Results concerning this resonance frequency as a function of the different physical and geometrical parameters of the antenna are presented. These obtained results correspond to the trained and tested data of the ANN model. A comparison with other results calculated from Chew's algorithm clearly shows the effectiveness of the proposed approach. The objective is to reduce the computational complexities, and thus to considerably reduce the computation time.