针对动力电池状态估计问题，引入分段遗忘因子函数，使遗忘因子随数据量动态调整，提出改进的带遗忘因子最小二乘算法(Improve Forgetting Factor Recursive Least Square，IFFRLS)对一阶Thevenin电池模型参数进行在线辨识，引入误差权重函数调整自适应拓展卡尔曼的过程噪声和测量噪声，提出改进的 自适应拓展卡尔曼算法(Improve Adapted Extended Kalman Filter，IAEKF)算法对电池 SOC 值进行估计，从而提出基于IFFRLS-IAEKF算法的在线辨识及电池荷电状态(State of Charge，SOC)估计方法。在动态应力测试(Dynamic Stress Test，DST)工况下仿真验证了所提方法具有更高的估计精度，为电动汽车的里程预测以及电池管理系统(Battery Management System，BMS)的电池管理提供理论支撑，具有重要的实用价值。
Online identification of IFFRLS-IAEKF algorithm and estimation of battery SOC
In response to the problem of power battery state estimation,a segmented forgetting factor function was introduced to dynamically adjust the forgetting factor with the amount of data.An improved Forgetting Factor Recursive Least Square(IFFRLS)algorithm was proposed to identify the parameters of the first-order Thevenin battery model online,and an error weight function was introduced to adjust and adaptively expand the Kalman process noise and measurement noise,proposed an improved Adaptive Extended Kalman Filter(IAEKF)algorithm to estimate the SOC value of batteries,thereby proposing an online identification and State of Charge(SOC)estimation method based on the IFFRLS-IAEKF algorithm.The simulation under the dynamic stress test(DST)condition verified that the proposed method had higher estimation accuracy,which provided theoretical support for the mileage prediction of electric vehicles and the battery management of battery management system(BMS),and had important practical value.
least squares methodAEKFbattery SOCDST working conditionparameter identificationonline estimation
万方数据
维普
