Statistical approach for continuous internal resistance estimation of lithium ion cells under dynamic loads

To meet the requirements of electrified vehicles, such as durability, driving range, fast charging and safety, an accurate and continuous estimation of the cell states on board is necessary. In this paper, a data-based method based on a statistical approach is derived and validated by experimental data. The purpose of the method is to realize a continuous resistance estimation of a lithium-ion cell under operation, with the demands of low computational effort and memory storage. The internal cell resistance of a cell is mainly influenced by the temperature and the State of Health (SoH) of the cell. Conversely, the data of the internal resistance can be used to extract conclusions regarding the temperature and the SoH. The presented algorithm uses the correlation of the voltage response to current changes. The variance of the current load resulting from the driving load correlates to the variance of the cell voltage caused by current variance. Under ideal conditions, this correlation factor corresponds to the internal resistance. The algorithm is recursively formulated and a quality factor is determined which evaluates the quality of the resistance estimation based on the available data. The method is investigated theoretically and validated by measurements on a lithium ion cell.