适用于具有显著电压平台区的LiFePO₄锂离子电池的鲁棒荷电状态估计

Abstract Accurate state of charge (SOC) estimation is critical for the safe and efficient operation of electric vehicles and energy storage systems. To address the challenges of reduced observability and noise sensitivity in voltage plateau regions of lithium iron phosphate (LiFePO 4 or LFP) batteries, this study proposes an adaptive robust extended Kalman filter (AREKF) with a dual-error collaborative mechanism for SOC estimation. First, the convergence condition, based on the identified open-circuit voltage (OCV) and estimated SOC, controls the activation and deactivation of SOC correction. Furthermore, by transforming the unmeasurable SOC condition into a measurable voltage condition based on state prediction and feedback errors, the correction window is expanded in the presence of SOC errors. Next, the error covariance of AREKF is adaptively updated by comparing the deviation between the calculated and theoretical voltage residual covariance, accelerating the convergence speed. Meanwhile, sliding window averaging and upper bounds on the error covariance are employed to enhance robustness. Finally, experimental validation demonstrates that the proposed method effectively suppresses measurement noise under dynamic conditions, exhibiting enhanced robustness in SOC estimation, particularly in the voltage plateau regions. Under multi-temperature, multi-noise, and disturbance testing, the steady-state estimation error remains within ±2 %, confirming the reliability of the proposed method for SOC estimation of LFP batteries.