基于Thevenin模型和自适应卡尔曼的SOC估算研究

精确的荷电状态(SOC)值在电池的应用开发中具有重要的意义。选择合适的滤波算法是精确估算的前提。由于扩展卡尔曼滤波(EKF)中噪声的给定值与实际工况下噪声的统计特性不符,导致估算精度低。为提高S0C估算精度,构建能准确反映锂电池工作特性的Thevenin电路模型。在此基础上,构建状态方程和观测方程,提出自适应卡尔曼滤波(AKF)算法。利用混合动力脉冲能力特性(HPPC)试验对模型参数进行辨识,通过MATLAB/Simulink建模仿真,分析锂电池分别在恒流放电和动态工况下S0C估算的精度。试验表明,Thevenin模型能够良好地表征锂电池的工作特性且能较好地进行S0C估算,参数辨识、恒流放电工况、动态工况下的最大误差分别控制在0.4%、0.2%、0.1%以内,验证了AKF应用于锂电池S0C估算的优越性。

Research on SOC Estimation Based on Thevenin Model and Adaptive Kalman

Accurate state of charge(SOC)value is of great significance in the application development of batteries.It is the premise of accurate estimation for choosing the appropriate filtering algorithm.Since the given value of noise in extended Kalman(EKF)does not match the statistical characteristics of noise under actual working conditions,the estimation is not accurate enough.To improve the accuracy of SOC estimation,a Thevenin circuit model that accurately reflects the operating characteristics of the lithium battery is constructed.Based on the construction of the state equation and the observation equation,an adaptive Kalman filter(AKF)algorithm is proposed.The hybrid pulse power characterization(HPPC)experiment is used to identify the model parameters.The MATLAB/Simulink modeling and simulation are used to analyze the accuracy of SOC estimation of lithium batteries under constant current discharge and dynamic conditions.Experiments show that Thevenin model can well characterize the working characteristics of lithium batteries and can better estimate SOC.The maximum error under parameter identification,constant current discharge and dynamic conditions is controlled within 0.4%,0.2%,and 0.1% respectively.It is verified for the superiority of AKF to SOC estimation of lithium battery.