Accurate battery models are essential for battery management systems to ensure reliable state estimation and performance prediction. However, the influence of capacity characterization conditions on equivalent circuit model (ECM) parameter identification remains insufficiently investigated. This study evaluates the effect of characterization C-rates on the parameter identification accuracy of a second-order Thevenin ECM using the Hybrid Pulse Power Characterization (HPPC) method. Three characterization conditions, namely C/20, C/5, and C/3, were applied to Panasonic NCR18650B lithium-ion cells to investigate the relationship between effective capacity variation, ECM parameters, and model accuracy. The results show that the effective capacity varied by less than ±2% acrossthe investigated C-rates. Model validation demonstrated that the C/20 and C/3 characterization conditions produced comparable prediction accuracies, with RMSE values of 0.0213 V and 0.0214 V, respectively, whereas the C/5 condition resulted in larger prediction errors. Residual analysis further confirmed that the C/3-based model exhibited uncertainty characteristics similar to those obtained under the conventional C/20 condition. These findings indicate that increasing the characterization rate from C/20 to C/3 does not significantly degrade ECM parameter identification accuracy while substantially reducing the experimental duration. Therefore, the C/3 characterization condition provides a favorable trade-off between model accuracy and testing efficiency, making it a promisingapproach for rapid battery characterization and battery management applications, particularly for lithium-ion battery systems used in LEO satellites.
Copyrights © 2026