基于联合算法的锂电池SOC与SOH协同在线预测

On-line prediction of lithium battery SOC and SOH based on joint algorithms

摘要:

以18650型锂电池为研究对象，建立双极化Thevenin(DP-Thevenin)等效电路模型描述其动静态特征。分别以恒流脉冲放电实验和带遗忘因子的递推最小二乘法完成电池电动势及模型参数的辨识；在Simulink中搭建等效电路模型，以脉冲电流作为激励进行验证，得出模型响应电压与实际端电压契合度较好，平均误差为1.836%；构建电池实验硬件电路，编写算法程序完成了锂电池实验系统的构建。最后，在随机测试工况下借助Matlab分析了基于联合算法的锂电池荷电状态(SOC)与健康状态(SOH)在预测精确度、错误初值时算法收敛性等方面的性能。实验结果表明，算法可精确估计出电池SOC和内阻大小，最大误差不超过3.5%；且在初值相差15%时，算法可在319 s内收敛至真值附近，鲁棒性较好

Abstract:

The equivalent circuit model of Dual Polarization Thevenin(DP-Thevenin) is established to describe the dynamic and static characteristics of type 18650 lithium battery. The open circuit voltage and model parameters are identified by constant current pulse discharge experiment and Recursive Least Squares method with Forgetting Factor(FFRLS). Then an equivalent circuit model is built in Simulink, and the impulse current is used as the excitation to verify the model. It is concluded that the response voltage of the model is in good agreement with the actual terminal voltage, with an average error of 1.836%. Next, the hardware circuit of battery experiment is constructed, and the algorithm program is compiled to complete the construction of lithium battery test system. Finally, the performance of State Of Charge(SOC) and State Of Health(SOH) of lithium batteries based on joint algorithm in predicting accuracy and convergence of the algorithm at wrong initial values is analyzed by means of Matlab under random test conditions. The experimental results show that the algorithm can accurately estimate the SOC and internal resistance of batteries, the maximum error is not more than 3.5%. When the initial value differs by 15%, the algorithm can converge to the true value within 319 s with good robustness.