Intelligent Modeling of Circular Microstrip Antenna for RF Energy Harvesting Applications

This paper introduces the application of Artificial Neural Network (ANN) modeling to a circular microstrip patch antenna for RFEH applications. Two distinct ANN models are proposed for the forward and inverse modeling of the antenna. The forward model takes the geometrical parameters as inputs and produces the reflection coefficient as the output. In contrast, the inverse model, adopts a reverse approach. The models exhibit root mean square error of 0.0120 and 0.0126 for training and 0.0145 and 0.0131 for testing, respectively. These low error rates indicate the models' effectiveness in accurately predicting the antenna's behavior. To further evaluate the performance of the proposed models, a correlation coefficient statistical approach is employed. This analysis provides additional insights into the reliability and correlation between the predicted values and the actual performance of the antenna.