Design of defected ground structures with artifical neural networks

In this thesis, it is aimed to create an artificial neural network model of a microstrip transmission line with Defected Ground structure. Thus by this mean, a simple, fast, efficient, low cost and accurate model for microwave applications such as matching networks, filter and antenna designs would be obtained. In order to have an accurate model, the training and test data for our model is obtained by using 3D EM simulation tools for different dielectric substrates with different sizes. The obtained data sets are used to ANN model of defected ground transmission lines. The simulation and ANN results are compared with measurement results at the end of the thesis. Also, an application based on ANN model of Defected Ground structure is presented for harmonic suppression of a class an amplifier.