Modeling the Battery Pack in an Electric Car Based on Real-Time Time-Domain Data

This paper illustrates circuital and rational functions models for the identification of the impedance of series-parallel connection of rechargeable batteries, using unevenly sampled time-domain data. Three circuital- and one-zero, two-pole rational function models are considered. Model parameters are identified using constrained minimization based on time-domain data. Approaches are validated using voltage and current signals sampled in a battery pack of a commercial car, collected during a long journey. It is shown that time-domain identification can be performed when the battery is functional and that model parameters exhibit a stationary behavior.