电池双滤波结构下模型参数与状态在线估计

在使用贝叶斯族估算电池SOC时由于电池初始参数及系统模型的不确定性,此类算法可能致使算法在运行中的某个时间点出现发散的情况。对此,本文采用改进型容积卡尔曼双滤波算法在线联合估计电池模型参数和SOC。同时,在对电池进行容量测试以及混合功率测试的基础上,使用对该算法电池进行评估。实验结果表明了电池状态估计的最大绝对误差2.95%,平均绝对误差位1.20%。通过与单一滤波结构的算法进行对比,说明了本双滤波结构在估算精度以及算法稳定型方面更优。

Dual Filtering Framework-based Estimation for Model Parameters and State of Charge on-line

When Bayes filter is used to estimate the SOC, it may diverge due to the uncertainty of the initial parameters and system model. In this paper, the double-scale improved CKF algorithm is proposed to estimate jointly the parameters of ECM and the SOC on-line. At the same time, based on the battery capacity test and hybrid power test, the algorithm is used to evaluate the battery. The experimental results show that the maximum absolute error of battery state estimation is 2.95%, and the average absolute error level is 1.20%. Compared with the single filter structure algorithm, the double filter structure is better in estimation accuracy and algorithm stability.