摘要
针对单一滤波算法对动力电池荷电状态(SOC)预测估计精度有限的问题,分析并建立了二阶RC网络等效电路模型,进行了离线参数辨识,并验证了辨识结果的准确性。以该模型为基础,运用无迹卡尔曼粒子滤波(UKPF)算法对动力电池SOC的动态模型状态进行预测估计,以带可变遗忘因子的递推最小二乘法(VFFRLS)对动态模型参数进行辨识,两者互为输入输出,实现UKPF-VFFRLS算法的联合估计。仿真实验结果表明:相比原有单一滤波算法,UKPF-VFFRLS联合估计算法使得SOC平均误差降低至0.74%,均方根误差(RMSE)低至0.009 9,提高了SOC的预测估计结果精度,从而提升了能源消耗预判能力和电池使用效率。
The estimation accuracy of power battery SOC is limited by single filtering algorithms,aiming at that problem,an equivalent circuit model of two-order RC network is analyzed and established,and offline parameter identification is carried out and the accuracy of the identification results are verified.Based on this model,using unscented Kalman particle filter(UKPF) algorithm to predict and estimate the dynamic model states of power battery,and introducing the variable forgetting factor recursive least squares(VFFRLS) algorithm for model dynamic parameter identification.The two parts are each other's input and output to realize the joint estimation of SOC by UKPF-VFFRLS algorithm.After simulation,the UKPF-VFFRLS joint estimation algorithm reduces the average SOC error to 0.74% and the RMSE to 0.009 9,which effectively improves the accuracy of SOC prediction and estimation results,thereby improving the ability to predict energy consumption and battery efficiency.
作者
许耀辉
张丽霞
刘大勇
常凤筠
XU Yaohui;ZHANG Lixia;LIU Dayong;CHANG Fengjun(School of Electronic and Information Engineering,University of Science and Technology Liaoning,Anshan Liaoning 114051,China;Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang Liaoning 110016,China;Liaoning Provincial Big Data Management Center(Liaoning Provincial Information Center),Shenyang Liaoning 110002,China)
出处
《电源技术》
CAS
北大核心
2023年第5期644-649,共6页
Chinese Journal of Power Sources
