ANN Model for Designing Stub Microstrip LowPass Filters

Using neural network technology, the design technology of the stubbed microstrip low-pass filter is given. The required length of the microstrip filter is used to obtain the correlation between inputs in the neural network. This article presents the design and analysis of stub microstrip low-pass filter with cut-off frequency of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1.48GHz</tex> with insertion loss (S21) of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3dB</tex> , and a return loss of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-2.59dB</tex> . Then, a neural network model was designed to find the difference in the irregularity (S-parameters) of the microstrip low-pass filter with a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1.48GHz</tex> range for <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</tex> -band applications. After the neural model of the microstrip low-pass filter was developed, its design was shown to be more feasible and precise as compared to simulation software. MATLAB programming language and industrial software, IE3D 14.1, are utilized to train the neural network for obtaining precise dimensions after performing simulation.