摘要
为了提高锂离子电池当前可用容量的预测精度,采用支持向量回归并对参数进行优化的算法。将样本中的部分边界向量作为支持向量,以平均放电电压、平均放电温度以及等压降放电时间序列作为输入,并结合优化算法对惩罚函数C和核宽度g两个参数进行优化,拟合出泛化性良好的容量估计方程。验证结果表明,采用遗传算法时,预测精度可高达99.6%。该方法无需推导具体的物理模型,对数据测量精度的要求较高,能够在各种锂离子电池中得到广泛的应用。
In order to improve the prediction accuracy of the current available capacity of lithium-ion batteries,this paper adopts the support vector regression algorithm and optimizes the parameters.By taking part of boundary vectors in samples as support vectors,taking average discharge voltage,average discharge temperature and constant voltage drop discharge time series as input,and combining optimization algorithm to optimize penalty function C and kernel width g,capacity estimation equation with good generalization was fitted.The results show that the prediction accuracy can reach 99.6%.This method does not need to deduce the specific physical model,and requires high data measurement accuracy.It can be widely used in a variety of lithium ion batteries.
作者
史永胜
马铭远
丁恩松
余强
李雷
SHI Yong-sheng;MA Ming-yuan;DING En-song;YU Qiang;LI Lei(College of Electrical and Information Engineering,Shaanxi University of Science&Technology,Xi’an Shaanxi 710021,China;Jiangsu Runyin Graphene Technology Co.,Ltd.,Yangzhou Jiangsu 225600,China)
出处
《电源技术》
CAS
北大核心
2019年第12期1996-2000,共5页
Chinese Journal of Power Sources
基金
国家自然科学基金项目(61871259)
关键词
锂离子电池容量
支持向量回归
遗传算法
lithium ion battery capacity
support vector regression
genetic algorithm