Classification method for Li-Ion batteries state of health estimation

In recent years more and more mobile applications are based on electric equipment and on batteries for providing the necessary energy to run them. The most common used battery technology today is the one based on Lithium-Ion. The process of estimating of the state of health for a battery is of high importance for any application based on electrochemical storage. In the last decade, various methods for estimating state of health for batteries have been proposed and studied. In this paper a classification method for estimating the deterioration of a lithium-ion battery is presented and tested on data provided on IEEE DataPort and from measurements carried out for three 18650 lithium-ion batteries. The algorithm applied shows encouraging results, showing a maximal error smaller than 5% for batteries in the interval of state of health between 80 and 100%, an interval where batteries are considered suitable for EVs. Another problem in any application that includes battery management systems, is the lack of documentation regarding the batteries. Besides the state of health estimation, in this paper a method for identifying the documentation of a battery is presented.