State of charge estimation of lithium-ion batteries based on multi-task learn and Cubature Kalman Filter

The State of Charge (SOC) of the battery is one of the core technologies in the Battery Management System (BMS), and its accurate estimation is crucial for the stable operation of the BMS. SOC estimation methods based on Equivalent Circuit Models (ECM) typically rely on the nonlinear relationship between the Open Circuit Voltage (OCV) and SOC. However, accurately obtaining this relationship still poses a challenge due to the complex and variable nature of environmental temperature and driving conditions. To address this issue, this paper proposes an online method for acquiring the OCV-SOC relationship based on Multi-Task Learning (MTL), while simultaneously identifying the parameters of the equivalent circuit model in real time. On this basis, the Cubature Kalman Filter (CKF) algorithm is combined to estimate the battery SOC online. The charge and discharge experiments were conducted under different conditions using 18,650 batteries, and the collected data was used for SOC estimation. The results show that the proposed MTL-CKF joint estimation method has high accuracy and robustness, with the mean absolute error and the root mean square error of SOC estimation remaining within 0.7%.