Battery voltage and state of power prediction based on an improved novel polarization voltage model

Abstract A reliable and accurate battery model is the basis of accurate prediction of battery voltage and state of power (SOP). Based on the electrochemical model of a battery, an improved novel polarization voltage (NPV) model based on current and time is developed in this study. The parameters of the NPV model can be identified with only small-batch primary data, and the accurate prediction of voltage and SOP at t-second (t > 0) under constant current (I ≠ 0) can be easily realized. The simulation results show that when 3C is used to charge the battery directly from state of charge = 0 to the cut-off voltage, the errors of terminal voltage and polarization voltage at the cut-off voltage are only 1.4% and 4.9%, respectively, and the average errors of them in the whole process are only 1.14% and 4%, respectively. The SOP prediction at t-second = 10 s with an overall error of around 5% can be realized by the NPV model. The NPV model is simple in structure and it is easy to identify model parameters. It is also very suitable for voltage simulation and SOP prediction, which can provide reliable support for charge and discharge safety control and energy management.