Determination of the resonance frequency of aparture coupled microstrip patch antenna with artificial neural networks

The aim of this study is to determine the resonance frequency of microstrip patch antenna with artificial neural networks (ANN). Having the highest bandwidth in its class, the aparture coupled microstrip patch antenna (ACMPA) has been studied on.In this study aparture coupled microstrip patch antenna (ACMPA) has been simulated by modelling with HFSS package software. In the result of simulation, resonance frequency values have been calculated, the required changes in antenna sizes have been made systematically. Each input parameter changes have been entered manually and output resonance frequency values have been recorded manually as well.The long period of time caused by the excessive calculation of simulation softwares used to obtain the desired antenna parameters has led to the search of new computer aided methods. Because it needs less information, responds to different problems, makes generalizations, has fast learning ability articial neural networks (ANN) has been chosen. In the training of ANNs using different learning methods on multi layer perceptrons, the performances of these methods have been compared.The results obtained from ANN models are in compliance with HFSS results. The advantages of the neural models used in this study are that the results are obtained within a short period of time and these results are correct.