基于数据统计特性的GS-SVM电池峰值功率预测模型

发布时间：2018-05-14 16:46

本文选题：动力电池 + 峰值功率　；参考：《电力自动化设备》2017年09期

【摘要】：以锰酸锂动力电池为研究对象,对电池处于不同温度和荷电状态下的情况进行10 s峰值功率测试,同时测量电池内阻。对实验测试得到的温度、荷电状态、内阻及峰值功率数据进行统计分析,包括测试变量间的相关程度评估和共线性检测,挖掘电池外特性参数与峰值功率数据间的统计关系。在此基础上,提出采用基于网格搜索的支持向量机(GS-SVM)建立电池的峰值功率预测模型。验证结果表明所提模型预测精度高,平均误差仅为3.65%;该模型训练时间短、响应速度快、操作性强,可以实现对动力电池峰值功率的快速估计,为电动汽车安全可靠运行提供有力保障。
[Abstract]:The peak power of lithium manganese oxide (LiMnO _ 4) battery was measured at different temperature and charge state for 10 s, and the internal resistance of the battery was measured at the same time. The data of temperature, charge state, internal resistance and peak power obtained from the experiment are analyzed, including the correlation degree evaluation among test variables and collinear detection, and the statistical relationship between the parameters of external characteristic of the battery and the peak power data is excavated. On the basis of this, the support vector machine (GS-SVM) based on grid search is used to establish the peak power prediction model of the battery. The results show that the proposed model has a high prediction accuracy and an average error of only 3.65. The model has the advantages of short training time, fast response speed and high maneuverability, which can realize the fast estimation of the peak power of the power battery. It can guarantee the safe and reliable operation of electric vehicle.
【作者单位】：北京交通大学国家能源主动配电网技术研发中心北京电动车辆协同创新中心;北京新能源汽车股份有限公司;国家电网山东省电力公司电力科学研究院;
【分类号】：TM912

【相似文献】