Model-based health condition monitoring method for multi-cell series-connected battery pack

Abstract

This paper has made efforts to investigate the two key parameters for indicating battery state-of-health. (1) A series dual adaptive extended Kalman filter algorithm has been proposed. (2) An online model-based capacity and resistance estimation scheme has been proposed, which estimates the uncertainty parameters of capacity and resistance to evaluate the health status of battery pack. The result indicates that the prediction inaccuracies of battery are less than 1%. It is helpful for monitoring the health status and safeguarding the safety application of battery system used in electric vehicles.