基于改进模型和无迹卡尔曼滤波的锂离子电池荷电状态估计

摘要

动力锂离子电池的荷电状态 (SOC) 估算是电池管理系统的核心与关键技术.传统的安时积分法会产生累积误差导致结果不收敛, 应用一阶Thevenin模型的各类卡尔曼滤波方法由于模型精度有限不能取得较好的估算结果.针对以上问题, 提出了以二阶Thevenin等效电路模型并结合动态元件参数, 采用无迹卡尔曼滤波 (UKF) 算法, 对状态方程改进的安时积分法结果进行修正, 提高了SOC估算的精度.在3种不同的工况下对电池进行SOC估算试验, 试验结果表明二阶Thevenin等效电路模型下UKF算法在SOC估算中可以快速收敛, 并取得较高的精度.%The estimation of the state of charge (SOC) of power lithium-ion battery is the core and key technology of battery management system.The traditional ampere-hour integration method will produce the cumulative error which causes the result non-convergent.The Kalman filter method with first-order Thevenin equivalent circuit model can not obtain a better estimation result because of the limited model precision.Based on the two-order Thevenin equivalent circuit model, this paper uses the unscented Kalman filter (UKF) algorithm to modify the results of the improved ampere-hour integration method and improves the accuracy of SOC estimation.The SOC estimation experiments of the battery were carried out under 3 different working conditions.The experimental results show that the UKF algorithm with the two-order Thevenin equivalent circuit model can quickly converge in SOC estimation and obtain a high accuracy.