工程应用EKF的动力锂电池SOC估算

Estimation of SOC of Lithium-ion Battery Using EKF

为了解决扩展卡尔曼滤波法在工程实践上估算电池荷电状态受到参数多、算法复杂等制约,而不适用于真实估算系统的问题,提出对动力电池各输入输出参数进行归纳的方法,建立一阶RC电池模型,以实测数据对应电池模型参数和扩展卡尔曼滤波估算中的变量,实现工程SOC估算。根据搭建的真实实验平台,并以CAN总线形式提供上来的电池数据,建立参数变量与SOC估算之间的联系,基于多信息输入和模型实现kalman滤波估算SOC。使用UBM4024动力锂电池组进行恒流、工况等放电实验,并设计PC上位机显示软件,对结果进行显示,结果验证工程条件下估算SOC值误差在5%以内。

Applying extended Kalman filter(EKF)algorithm in estimating SOC of batteries in engineering practice has limitations as multi-parameter and algorithm complexity.To solve this problem,an inductive method based on the input and output data of lithium battery is put forward.And it achieves SOC estimation by establishing a first-order RC battery model and substituting variables in battery model and EKF estimation with measured data.An experimental platform is built.The battery data is obtained using CAN bus to set up contact between variables and SOC estimation.Discharge experiments using a UBM4024 lithium battery in constant-current working conditions is conducted and the results are also displayed on a software interface in PC.The EKF SOC estimation based on multi-information input and battery model has been achieved with an error less than 5% under the project conditions.