Optimization of microstrip dual mode resonator filter structure modeling using artifical neural network applications

Abstract

Artificial Neural Network computational modules have recently gained recognition RF and microwave modeling. This situation is very important for computer aided tuning of microwave components. Neural networks can be trained to learn the behavior of passive/active components/circuits. This paper presents a design approach for a microstrip dual mode filter by using the artificial neural network modeling (ANN) technique. First to develop new models available in the programs a simple modeling technique were used. To these end important dimensions of the filter layout are used to capture critical input output relationships. Once fully developed, the ANN model has been shown to be as accurate as an EM simulator and much more efficient results in the filter optimization.