关键词: 电池管理系统, 荷电状态和健康状态联合在线估计, 鲁棒无迹卡尔曼滤波, H无穷滤波, 在线参数辨识

Abstract: An accurate estimation of State of Charge (SOC) and State of Health (SOH) algorithm for lithium-ion batteries is crucial for Battery Management Systems (BMS). Traditional Unscented Kalman Filter (UKF) algorithm suffers from limited accuracy under complex operating conditions due to noise and model parameter uncertainties. To address these limitations, a Dual-time-scale Robust Unscented Kalman Filter (DRUKF) algorithm based on a second-order RC equivalent circuit model and integrated with online parameter identification is proposed for joint online estimation of SOC and SOH. The DRUKF algorithm exploits the slowly varying characteristics of SOH to achieve decoupled estimation of SOC and SOH. By integrating H-infinity filtering with Robust Unscented Kalman Filter (RUKF) algorithm for online parameter identification,the algorithm can dynamically update both model and filtering parameters. The H-infinity filtering reduces the algorithm′s dependence on noise statistical properties,while the RUKF algorithm enhances robustness against model parameter deviations. Experimental results demonstrate that the SOC estimation error of the DRUKF algorithm is significantly lower than that of the Extended Kalman Filter (EKF),UKF, and RUKF algorithms based on offline parameter identification. These experiments validate that the DRUKF algorithm achieves high accuracy and strong robustness under complex operating conditions, thereby providing a reliable and practical solution for BMS applications.

Key words: Battery Management Systems (BMS), joint online estimation of State of Charge (SOC) and State of Health (SOH), Robust Unscented Kalman Filter (RUKF), H-infinity filtering, online parameter identification