Unscented Kalman Filter based Coestimation of SoC and SoHR in Lithium Battery with Hysteresis

Lithium iron phosphate batteries are the most promising energy storage systems for automotive applications thank to their low cost, good energy density and high safety. However, they exhibit a pronounced hysteresis between charge and discharge voltage curves that significantly complicates the estimation of the battery States of Charge and Health. A state estimation algorithm based on the Unscented Kalman filter is developed in this work to estimate the State of Charge and the resistance State of Health of a Lithium-iron-phosphate battery. The hysteresis is modeled by a first-order charge relaxation equation. Moreover, a custom simulation platform is developed to test the estimation algorithm in automotive applications obtaining a State of Charge estimation error lower than 1 % and a voltage estimation RMS error of only 1.94 mV when the battery cell is exercised with the UDDS load profile.