Battery as energy storage equipment plays an important role in social life and industrial production. Lithium-ion batteries have already been widely used in different power demand occasions because of excellent performance. The battery management system (BMS) is the critical part of lithium-ion battery applications. Accurately estimating battery status is the core function of BMS, which relies on the battery model and parameters of model identification. In order to obtain the dynamical parameters precisely online, this paper modifies Shepherd mathematical battery model and decouples connectivity of model parameters to facilitate iterative calculation. Then use least squares method to track the state change of lithium-ion battery in real time through dynamic parameters estimation. The validity is demonstrated by simulation and experiment.