The Effects of Test Profile on Lithium-ion Battery Equivalent-Circuit Model Parameterization Accuracy

Abstract

Accurate modelling of battery cells is crucial for the safety and longevity of the battery system. The equivalent-circuit battery model (ECM) is widely used because it provides good accuracy at a relatively low computational cost. The accuracy of an ECM depends primarily on the model parameters, which can be identified using optimization algorithms based on experimental data. This study investigates the effect of test profiles on the accuracy of the ECM by comparing (1) pulse tests of various lengths (2) two identification methods - direct optimization method and analytical method and (3) identification with pulse and drive cycle tests. The results suggest that optimization with an application-specific test profile (drive cycle tests for example) can provide the best accuracy. Parameters identified from the pulse test with short rests using the analytical method provided comparable accuracy, suggesting that the commonly used 30 to 120 minutes rest lengths may be unnecessary. Finally, to obtain a continuous relationship between the open-circuit voltage (OCV) and the cell’s state of charge (SOC), a polynomial is fitted to the OCV curve. Polynomials with orders from 5th to 21st are tested and it was found that 11th order polynomial provided a good compromise between the complexity and model accuracy.