State-of-Charge Estimation for Lithium Iron Phosphate Batteries Using A Multi-Model Filter

The accuracy of State-of-Charge (SOC) estimation is a key concern in the application of Lithium Iron Phosphate (LFP) batteries. In this paper, a novel SOC estimation method is proposed based on the Interactive Multi-Model Unscented Kalman Filter (IMM-UKF) algorithm. Four models are set up to represent different working conditions of LFP batteries, and thus, the requirement for single-model accuracy is lowered. UKF is used to estimate the states of each single model which are then interactively mixed to generate the global estimated SOC. The nonlinearities in battery models are well dealt with by using UKF. Dynamic tuning techniques help improve the accuracy of SOC estimation. Experiment results show satisfactory SOC estimation accuracy. 24%~78% improvements on SOC estimation performance can be observed under various error metrics and working conditions, compared to the interactive multi-model extended Kalman filter method.