基于分数阶联合卡尔曼滤波的磷酸铁锂电池简化阻抗谱模型参数在线估计

摘要

Battery modeling and online battery model parameter estimation are the key technologies of EV battery management system. Based on the battery simplified electrochemical impedance spectroscopy which contains a fractional component, this paper establishes the state transition and systematic observation equations for the nonlinear system of LiFePO4 secondary battery. Then, the diffusion polarization voltage and model parameters are estimated online with the fractional joint Kalman filter (FJKF). The experimental results show that, this model can reflect the dynamic characteristics very well, and FJKF parameter estimation algorithm can maintain good accuracy. Meanwhile, the method is suitable for a variety of load conditions. The model parameters obtained by this algorithm have good stability.%电池特性建模及模型参数在线估计是电动汽车电池管理系统的关键技术，以磷酸铁锂电池这一非线性系统为研究对象，以包含分数阶元件的简化电池电化学阻抗谱模型为基础，建立了该模型的状态转移方程和系统观测方程，运用分数阶联合卡尔曼滤波器（FJKF）对该模型的扩散极化电压和模型参数进行了在线估计。试验结果表明，该模型能较好地表征磷酸铁锂电池的动态特性，分数阶联合卡尔曼滤波算法在参数估计过程中能够保持很好的精度，同时该方法对多种测试工况都有较好的适用性，算法估计得到的模型参数值具有较好的稳定性。