| Title |
Analysis on Hysteresis Characteristics and Extended Kalman Filter-Based State of Charge Estimation of LFP Batteries |
| Authors |
Eun-Jin Kang ; Min-Hyeok Kim ; Jong-Hoon Kim |
| DOI |
https://doi.org/10.6113/TKPE.2024.29.4.300 |
| Keywords |
Lithium-iron phosphate battery; Extended Kalman filter; Recursive least squares; State-of-Charge; Hysteresis characteristics |
| Abstract |
Lithium-iron phosphate (LFP) batteries have high power and energy density, as well as long cycle life characteristics. LFP batteries also have the advantage of being safer than nickel?cobalt?manganese (NCM) batteries, which are widely commercialized domestically. On the other hand, LFP batteries have different open-circuit voltage (OCV) characteristics and hysteresis characteristics compared with NCM batteries, making estimating their state-of-charge (SOC) using SOC?OCV tables difficult. This paper therefore proposes a method for estimating the state of charge through the design of an extended Kalman filter and an optimal voltage model based on hysteresis characteristics. The Thevenin electrically equivalent circuit model was designed by deriving parameters for each state of charge using the recursive least squares method. To compare the state of charge estimation with consideration of hysteresis characteristics, a hysteresis model was incorporated for estimating the state of charge based on driving profiles. The validation results of the proposed method demonstrate model error performance of less than 1.9%. |