自适应无迹卡尔曼滤波动力电池的SOC估计
SOC estimation of power battery based on AUKF

摘要 无迹卡尔曼滤波法(Unscented-Kalman Filter,UKF)在估计动力电池的剩余容量(State of Charge,SOC )时，由于系统噪声的不确定，可能导致算法不收敛，而且算法的估计性能受模型精度的影响，为此采用自适应无迹卡尔曼滤波法(Adaptive-UKF,AUKF)动态估计电动汽车动力电池的SOC.建立了适用于SOC估计的电池模型，辨识相应的电池模型的参数并进行验证，将AUKF应用到该模型，在未知干扰噪声环境下，在线估计电池的SOC.试验仿真结果表明：UKF算法的估计误差在-0.04~0.06之间跳动，而AUKF算法的估计误差平稳的保持在0.05以内，实时修正微小的模型误差带来的SOC估计误差.
Abstract：The UKF method can be used to estimate the SOC of power battery, however, the uncertainty of the system noise may cause that the algorithm does not converge, and the estimation performance of the algorithm is affected by the accuracy of the model. An AUKF is used to estimate the dynamic SOC of an electric vehicle. At first, an equivalent circuit model appropriate for SOC estimation is built and the corresponding parameters of the battery model are identified. The AUKF is used in this model for online estimation of battery SOC in unknown noise environment. Experimental results show that the estimation error of UKF algorithm is beating between -0.04~0.06, while the estimation error of AUKF algorithm is kept within 0.05 and the SOC estimation error is corrected in real time.