ANN Based Design of a Rapid Battery Charger

Abstract

This paper presents an application of ANDEng, an Artificial Neural Network Design Automation tool for the design of a rapid Nickel Cadmium battery charger. Rapid charging needs high charge current of the order of 8C where C represents the capacity of the battery. The battery charger needs an intelligent strategy as the charging has to be controlled to avoid damage to the battery due to rise in temperature caused by high charging currents. The temperature control assumes significance as the rise in temperature is exponential after initial charge. The paper also presents the analog implementation of neurons with log sigmoid and linear activation functions using discrete components.