Relax, Estimate, and Track: a Simple Battery State-of-charge and State-of-health Estimation Method (2408.01127v3)

Abstract: Battery management is a critical component of ubiquitous battery-powered energy systems, in which battery state-of-charge (SOC) and state-of-health (SOH) estimations are of crucial importance. Conventional SOC and SOH estimation methods, especially model-based methods, often lack accurate modeling of the open circuit voltage (OCV), have relatively high computational complexity, and lack theoretical analysis. This study introduces a simple SOC and SOH estimation method that overcomes all these weaknesses. The key idea of the proposed method is to momentarily set the cell's current to zero for a few minutes during the charging, perform SOC and SOH estimation based on the measured data, and continue tracking the cell's SOC afterward. The method is based on rigorous theoretical analysis, requires no hyperparameter fine-tuning, and is hundreds of times faster than conventional model-based methods. The method is validated on six batteries charged at different C rates and temperatures, realizing fast and accurate estimations under various conditions, with a SOH root mean square error (RMSE) of around 3% and a SOC RMSE of around 1.5%. The data and codes are available at https://berkeley.box.com/s/jz1w6po2iqzzfy7irxd9ok47ku3tr86j.