Prediction of Energy Storage Capacitor Values Based on Neural Networks. (Case of a Planar Capacitor)

Abstract

Energy storage is a very important operation in continuously operating systems, such as telecommunications systems, embedded systems and power systems. Energy storage can be performed by various means such as batteries and super capacitors. In our work, we used neural networks to determine the capacitance values C of the planar capacitors as a function of the relative permittivity εr, the distance d and the dimensioning (Width and Length) of the capacitor plates and as a function of the maximum desired charge Qmax. The results of simulation will be better and more satisfying if the databases are richer and good.