基于自适应电池模型的SOC加权在线估计

Weighted Online Estimation of SOC Based on Adaptive Battery Model

针对传统锂电池模型在SOC估计上自适应能力差,以及单一SOC估计算法局部估计精度低的问题,提出了一种利用多新息随机梯度算法进行模型参数在线辨识及SOC(荷电状态)加权在线估计方法。锂离子电池模型参数在线辨识结果实时更新,实现锂离子电池模型自适应。针对SOC的估计,提出了基于PI(比例积分)调节器的开路电压法并结合安时积分法的加权在线估计方法,权值依据当前SOC–E_0分段线性曲线斜率动态更新,解决了开路电压法平台区估计误差大、安时积分法难以确定初值和误差累积以及两种方法难以在线估计的问题。从理论分析上论证了方法的可行性,并从MATLAB仿真结果上验证了所提方法具有较高的估计精度。

Aiming at the problem that the traditional lithium battery model has poor adaptive ability in SOC estimation and the local estimation accuracy of the single SOC estimation algorithm is low, online identification of model parameters with Multi-Innovation Stochastic Gradient algorithm and method of weighted online estimation of SOC was proposed. The results of lithium-ion battery model parameters online identification updated in real time, realizing lithium-ion battery model adaptive. Aiming at the estimation of SOC, a method of weighted online estimation method based on the PI （proportion and integral） regulator＇s OCV combined with Ah （an integral method） was proposed. Weights updated in real time according to SOC-Eo piecewise linear curve slope. The proposed method solved that the estimation error of SOC with OCV is large and Ah method is difficult to determine the initial value and cumulative error Besides, the proposed method also solved the problem of online estimation of above two method. The feasibility of the method is demonstrated from the theoretical analysis, and the results of MATLAB simulation show that the proposed method has high estimation accuracy.