Comparative evaluation of LiFePO4 cell SOC estimation performance with ECM structure and noise model/data rejection in the EKF for transportation application

Abstract

This study presents a comparison of several methods to improve state-of-charge (SOC) estimation performance using the extended Kalman Filter (EKF) algorithm for a commercial a lithium iron phosphate (LiFePO 4 ) cell. Firstly, this work attempts to show the comparison of SOC performance according to the number of RC-ladder. Secondly, this work shows a comparison of SOC estimation with and without minor loop to overcome the difference between charging open-circuit voltage (OCV) and discharging OCV. The SOC performance with and without noise model and data rejection in the EKF is finally compared.

Comparative Analysis of SOC Estimation using EECM and NST in Rechargeable LiCoO2/LiFePO4/LiNiMnCoO2Cells

“…18. In contrast to the major loop in the EECM, the minor loop application greatly improves the accuracy of OCV in the EECM, even under the OCV hysteresis effect of the LiFePO 4 cell [23].…”

“…The most important thing to be considered in this approach is that the LiFePO 4 cell has unusual open-circuit voltage (OCV) characteristics caused by the hysteresis effect. Therefore, it is so difficult to have an accurate SOC estimation using a conventional OCV, namely major loop [20][21][22][23]. The accuracy of the EECM can be more increased by developing some minor loops in the conventional charging/discharging OCV curves (major loop).…”