Analysis of two Equivalent Circuit Models for State of Charge Estimation using Kalman Filters

Abstract

Battery State of Charge estimation is critical to improve performance, lifetime, and safety of batteries used in agricultural robotic systems. This paper analyzes two different Equivalent Circuit Models that can be used for State of Charge estimation of batteries. We analyze the RC model proposed by the National Renewable Energy Laboratory and the 2RC Thevenin model. The main objective of this study is to determine the advantages and limits of equivalent circuit models for online State of Charge tracking. We performed, for each model, an estimation of the parameters offline. Also, the Recursive Least Square technique was implemented for one of the models to analyze the effect of the online estimation in battery modeling. Then, we implemented and analyzed two Kalman Filters for State of Charge estimation: Linear Kalman Filter and Unscented Kalman Filter. Finally, we provide simulation results of models implemented using MATLAB that we evaluated using experimental data from tests performed on a Panasonic 18650PF Li-ion Battery.