Neuromodeling stop band properties of Koch Island patch elements for FSS filter design

This work investigates the use of Koch Island patch elements to design frequency selective surface (FSS) band stop filters that reflect X‐ or Ku‐band signals. To control the FSS resonant frequency and −10 dB bandwidth, we modify the shape of Koch Island patch elements adjusting the fractal parameters, iteration factor, and iteration number. The electromagnetic (EM) behavior is described as a function of the element shapes and the permittivity of the dielectric substrate. A FSS parametric analysis is accomplished through the use of the Ansoft Designer™ software. The obtained EM‐data is utilized for training a multilayer perceptrons (MLP) artificial neural network to achieve the FSS modeling task. Experimental and simulated results are obtained to validate the developed MLP model.