平方根无迹卡尔曼滤波作球面变换的SOC估计  

State-of-charge estimation based on square root unscented Kalman filter with spherical transformation

作　　者：何俊儒[1] 王洪诚[2] 杨欣荣[1] 王蕾[1] 

机构地区：[1]西南石油大学机电工程学院,四川成都610500 [2]西南石油大学电气信息学院,四川成都610500

出　　处：《电源技术》2018年第1期114-118,共5页Chinese Journal of Power Sources

摘　　要：针对现有的电池荷电状态(SOC)估计方法存在计算推导过程复杂以及线性化精度低的缺点,提出了一种新的基于平方根无迹卡尔曼滤波在单位超球体中作球面变换的锂电池SOC估计方法。这种方法无需对非线性模型线性化且与传统的无迹卡尔曼滤波相比,通过球面变换得到的Sigma点也更少,从而降低了计算要求。修正了电池的二阶等效电路模型,然后给出了所提出估计方法的具体步骤。最后,通过实验对估计方法进行了验证,分析了所提出的方法在SOC估计精度和鲁棒性方面的性能。实验表明,所提出的估计方法能顺利地完成电池SOC的精确估计,估计误差最大仅为4.98%,估计精度受参数变化影响小,具有一定的鲁棒性。For the complex calculation and derivation and low linearization accuracy of exiting battery state-of-charge(SOC) estimation, a new SOC estimation of lithium-ion battery based on square root unscented Kalman filter with spherical transformation in unit hyper sphere was proposed. The method dose not require linearization for nonlinear model, and compared with traditional unscented Kalman filter, the Sigma points obtained by spherical transformation is less, which reduces computational requirements. The two order equivalent circuit model of the battery was fixed.The detail procedures and algorithms of the proposed estimation method were given. The method was verified by experiments. The performance of the proposed method in SOC estimation accuracy and robustness was analyzed.The experiments show that the accurate SOC estimation of battery can be successfully accomplished by the proposed method, and the maximum estimation error is only 4.98%. The estimation accuracy affected by parameters variation is small with a certain robustness.

关 键 词：锂电池 二阶等效电路模型 SOC 平方根无迹卡尔曼滤波 球面变换 

分 类 号：TM912.9[电气工程—电力电子与电力传动]