Feed forward modelling of nature-inspired planar antenna using artificial neural network (ANN)

This article proposes the artificial neural network (ANN) modeling of a nature-inspired planar antenna using the Adaptive Moment Estimation (Adam) optimizer. The ANN model will predict the values of output performance parameters, such as return loss, Gain, VSWR, and resonant frequency, with respect to variations in the input dimension of the nature-inspired (flying insect) patch structure. The predicted output performance parameters are compared with those simulated by the CST software for different input testing design parameter samples of the flying insect patch structure. Simulation is conducted using CST software, considering a FR4 substrate with a dielectric constant of 4.4 for the antenna.