In this paper, a real time estimate of a battery model is proposed. The existing impedance model is accurate in predicting the behavior of a battery. However, it is complex and not applicable for real time applications. Also, it requires much time and cost to obtain the parameters of the model. A simple lumped circuit model is presented. The parameters and states (overvoltages) of the battery are estimated with a filtering technique and system identification theory. It is verified through simulations and experiments that the proposed method is in good agreement with the dynamic characteristics of the battery, regardless of the conditions of the battery, and the estimated overvoltages are given physical meanings through a comparison with an impedance model with complex RC ladder circuits