基于K近邻回归的锂离子电池健康状态估计  

State of health estimation of lithium-ion battery based on K nearest neighbour regression and IC curve

作　　者：孙广明 贾新羽[3] 陈良亮 SUN Guangming;JIA Xinyu;CHEN Liangliang(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing Jiangsu 211106,China;NARI Technology Co.,Ltd.,Nanjing Jiangsu 211106,China;National Active Distribution Network Technology Research Center(NANTEC),Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]南瑞集团(国网电力科学研究院)有限公司,江苏南京211106 [2]国电南瑞科技股份有限公司,江苏南京211106 [3]北京交通大学国家能源主动配电网研究中心,北京100044

出　　处：《电源技术》2022年第8期872-875,共4页Chinese Journal of Power Sources

基　　金：国家重点研发计划科技助力经济2020重点专项(2020YFB0100300ZL)。

摘　　要：锂离子电池健康状态的降低伴随着电池内部的衰退,并会引发电池鼓包或者短路等安全问题,在锂离子电池充电过程中准确地估计电池健康状态能够为电池的实际使用与充电管理提供重要的参考。从三元锂离子电池充电过程中的容量增量曲线上提取表征三元锂离子电池的健康状态参数,然后利用K近邻算法对三元锂离子电池的健康状态进行估计。利用K近邻回归的机器学习方法拟合了电池衰退轨迹,基于K近邻的电池SOH估计的决定系数R2>0.98。The reduction of the status of health of lithium-ion batteries will cause the loss of the material and Li-ion.The internal degradation of batteries will lead to bulges or internal short circuit issues.Therefore,accurate estimation of the status of health of lithium-ion batteries during the charging process will provide meaningful information for the charge management and utilization of batteries.This paper extracted the features that reflected the status of lithiumion batteries'health from the incremental capacity(IC)curve during the charging process.Then the K nearest neighbour algorithm was used to estimate the state of health(SOH)of the lithium-ion battery.The ageing curves of lithium-ion batteries were fitted well by the K nearest neighbour regression method.The coefficient of determination of the SOH estimation based on K nearest neighbour regression is larger than 0.98.

关 键 词：锂离子电池 机器学习 健康状态估计 充电容量增量曲线 K近邻回归 

分 类 号：TM912.9[电气工程—电力电子与电力传动]