A physics-based fractional-order equivalent circuit model for time and frequency-domain applications in lithium-ion batteries

Abstract

Equivalent circuit models (ECMs) remain the most popular choice for online applications in lithium-ion batteries because of their simpler parameterization and lower computational requirements in comparison to electrochemical models. Nevertheless, standard ECMs lack physical insight and fail to accurately reproduce cell behavior under a wide range of operating conditions. For this reason, the development of physics-informed ECMs becomes essential so as to provide a better description of the physical processes while maintaining a reduced computational complexity. In this article, we propose a novel physics-based ECM derived directly from an electrochemical model, so that there is a clear correlation between circuit states and internal battery states, as well as circuit and physical parameters. The proposed model yields an RMS error below 1.46 mV for cell voltage, 0.28% for the surface concentration in the active material particles, 0.6% for the electrode-averaged electrolyte concentration and 0.32 mV for the charge-transfer overpotentials. Another key feature of this model is the relationship between circuit parameters and those identified in frequency-domain tests, which allows us to characterize and validate the model experimentally. We understand that the presented model constitutes an alternative to standard ECMs as well as electrochemical models as it combines advantageous characteristics from both of them.