基于强跟踪卡尔曼滤波的电池SOC估计

针对锂电池模型不准确和状态突变导致SOC估计精度不佳的问题,提出了引入时变渐消因子的强跟踪卡尔曼滤波算法.以HPPC试验方法辨识了锂电池的等效二阶RC模型,对比分析了现有的扩展卡尔曼滤波原理及提出的强跟踪卡尔曼滤波算法.通过结合强跟踪原理和卡尔曼滤波算法并引入时变渐消因子,提出的方法能够强制估计残差保持正交特性,并保证残差满足高斯白噪声特性.仿真验证表明,与扩展卡尔曼滤波原理相比,在模型不准确和状态突变的情况下,强跟踪卡尔曼滤波算法具有更高的估计精度,估计误差低于2.5%,提高了近45%.

Battery SOC estimation based on strong tracking Kalman filtering

In order to solve the poor accuracy problem of SOC estimation caused by the inaccurate lithium-ion battery model and sudden state changing, a strong tracking Kalman filtering algorithm with time varying fading factor was proposed. The equivalent two-order RC model for lithium-ion battery was identified by the HPPC test method, and the existing extended Kalman filtering principle and the proposed strong tracking Kalman filtering algorithm were compared and analyzed. Through combining the strong tracking principle and Kalman filtering algorithm as well as introducing the time varying fading factor, the proposed method can imperatively estimate the residual error to maintain the orthogonality and make the residual error satisfy the Gauss white noise characteristics. The simulation verification shows that compared with the extended Kalman filtering principle, the proposed strong tracking Kalman filtering algorithm has higher estimation accuracy under the condition of both inaccurate model and sudden state changing, and the estimation error is less than 2.5%, where the accuracy increases nearly by 45%.