The prediction of SOC of lithium batteries and varied pulse charge

Abstract

Improved RBF neural network arithmetic is mainly characterized by using TI's Impedance Track TM technology for reference which predict the status of charge (SOC) of lithium battery, and in accordance with the chemistry characteristics of lithium batteries, use varied pulse charge method for their rapid and efficient charging. The results show that the SOC which is predicted by improved RBF arithmetic can meet the performance target under the C++ compiler environment and adopting the varied pulse charge have shorten the charging time by 20%, this method is suitable for rapid charging system.