基于鲁棒H∞滤波的蓄电池荷电状态估计

针对蓄电池系统的荷电状态(SOC)受蓄电池材料及加工制作、工作温度、充放电大小及频率等因素的影响,是一个典型的非线性时变系统,相应的状态估计模型在测量过程中存在噪声干扰引起模型参数不确定性的特征。以安时法为基础,建立SOC的状态方程并应用鲁棒H∞滤波算法预测SOC估计值。仿真研究表明,提出的鲁棒H∞滤波算法在有色噪声干扰下比卡尔曼滤波(Kalman filter)有更好的估计精度;在白噪声情况下,鲁棒H∞滤波算法可通过调节其参数达到和卡尔曼滤波器相同的估计精度。

Estimation of State of Charge for Storage Battery Based on Robust H∞Filter

State of charge (SOC) of batteries, determined by materials and manufacture, operating temperature, charge/discharge current and frequencies, is a typical nonlinear time- varying system, and the corresponding model for state estimation is characterized by parameter uncertainty because of noise involved in measuring process. Based on Ampere hour method, state equations of SOC are set up and robust H_∞ algorithm is applied to estimate SOC. Simulation shows that the proposed robust H_∞ algorithm in the case of color noise has better performance of estimation accuracy than Kalman algorithm, and in the case of white noise has the same estimation accuracy with Kalman filter by parameter regulating.