Аңдатпа
Radio Frequency (RF) switches are one of the key drivers in the modern design of Frequency-Reconfigurable Circuits (FRCs). Recently, fully printed Vanadium Dioxide (VO2) based RF switches have been proposed for the state-of-the-art FRCs. However, VO2switch is still an evolving technology and needs behavioral models for the possible inclusion in computer aided design tools. This paper explores and develops Artificial Neural Network (ANN) and Grey Wolf Optimizer (GWO) aided ANN based behavioral models for the VO2switch. Then, the models are thoroughly evaluated on common metrics namely mean squared error, mean absolute error, coefficient of determination, prediction ability on unseen data, training and simulation time, parameters’ tuning time and complexity of the models.