摘要
蓄电池在直流微网中广泛应用,在使用过程中,需要对蓄电池的荷电量进行在线预测,即进行SOC估计,对蓄电池的模型以及参数辨识有较高的要求.由于蓄电池同时具有电特性与化学特性,建立精确的模型比较复杂,本文对蓄电池常用的5种模型进行了对比,并基于混合动力脉冲能力特性HPPC实验结果分别对Thevenin模型和GNL模型进行参数辨识,通过MATLAB仿真对2种模型的准确性进行了对比,在应用扩展卡尔曼滤波法进行SOC估计时,分别对2种模型参数得到的估计结果进行了比较验证.
Batteries are widely used in DC Microgrid.In the process of use,the SOC estimation of the battery must be predicted online,which needs high requirements of the battery model and the parameters identification.Because both of the electrical and chemical properties of the battery,the building of the accurate model of battery is complex.In this paper,five models are introduced and compared,the parameters of Thevenin and GNL models are identificated based on the Hybrid Pulse Power Characteristic HPPC experiment.The accuracy of the two models are compared through the matlab simulation.Finally the estimation results of two models parameters are compared and verified by using the EKF（the extended kalman filtering technique）estimation of SOC.
作者
刘欣博
边亚伟
王慧娴
LIU Xinbo;BIAN Yawei;WANG Huixian(Inverter Technique of Engineering Technology Research Center of Beijing, 100144, Beijing, China;Collaborative Innovation Center of Key Power Energy-Saving Technique of Beijing, 100144, Beijing, China;Collaborative Innovation Center of Electric Vehicles of Beijing, 100144, Beijing, China)
出处
《北方工业大学学报》
2018年第2期27-35,73,共10页
Journal of North China University of Technology
基金
北京市市属高校创新能力提升计划(PXM2013_014212_000069_00062484)
北方工业大学2016年优秀青年教师培养计划(XN072-013)
