Analysis of ECM Battery Modeling Techniques for Different Battery Types

The paper presents a comparative study of two methods for identifying equivalent circuit models (ECM) parameters for battery modeling. In order to comprehensively compare the methods, the electrical behavior of a lithium titanate (LTO), a lithium iron phosphate (LFP), a lithium nickel manganese cobalt oxides (NMC), and a nickel cobalt aluminium (NCA) battery are analysed. A second-order ECM is used to model batteries, while two different theoretical methods are applied to identify the ECM's parameters. The first method finds a fitted function for battery performance based on time response of ECM dynamics, while the second method uses a subspace system identification approach. The models are compared by calculating the errors between simulated data and measurement data based on the discharge pulse power characterization (DPPC) test. Simulation results show that the second method is more accurate in batteries modeling, while the values of the identified parameters are more stable with respect to changes in state of charge (SOC).