基于恒压充电数据的锂离子电池SOH估计  

SOH estimation of Li-ion battery based on constant voltage charging data

作　　者：杨驹丰 李哲 王振 邬明宇 马迷娜 栗欢欢[1] YANG Jufeng;LI Zhe;WANG Zhen;WU Mingyu;MA Mina;LI Huanhuan(Automotive Engineering Research Institute,Jiangsu University,Zhenjiang,Jiangsu 212013,China;School of Mechanical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China;School of Safety Engineering and Emergency Management,Shijiazhuang Tiedao University,Shijiazhuang,Hebei 050043,China)

机构地区：[1]江苏大学汽车工程研究院,江苏镇江212013 [2]上海交通大学机械与动力工程学院,上海200240 [3]石家庄铁道大学安全工程与应急管理学院,河北石家庄050043

出　　处：《电池》2024年第6期815-820,共6页Battery Bimonthly

基　　金：国家自然科学基金(52307244);国家资助博士后研究人员计划(GZC20231585);中国博士后科学基金第74批面上项目(2023M742255);江苏省自然科学基金(BK20210773);河北省自然科学基金(E2022210060);智能绿色车辆与交通全国重点实验室开放基金(KFY2408)。

摘　　要：准确估算健康状态(SOH),对锂离子电池的安全运行至关重要。基于恒压(CV)充电工况中的电流数据,提取电流曲线差异性参数作为健康特征(HI),并结合灰狼优化(GWO)-支持向量机回归(SVR)算法,构建HI与SOH的映射关系,获得SOH估计模型。基于两组公开电池测试数据集的验证表明,在完整及非完整CV充电工况下,所提方法的SOH估计均方根误差均低于2%。基于GWO-SVR、SVR和高斯过程回归等算法的电池SOH估计误差表明,所提方法的综合性能较好。Accurate estimation of state of health(SOH) is crucial for the safe operation of the Li-ion battery.Based on the current data under the constant voltage(CV) charging scenario,the difference parameters of the current curves are extracted as health indicators(HI).In order to obtain the SOH estimation model,the gray wolf optimization(GWO) and the support vector regression(SVR) algorithms are combined to construct the mapping relation between HI and SOH.The validation based on two public battery test datasets demonstrates that under both the complete and the incomplete CV charging scenarios,the root-mean-square errors of the SOH estimation by the proposed method are overall less than 2%.The SOH estimation errors are compared with algorithms including GWO-SVR,SVR and Gaussian process regression.it indicates that the proposed method has better comprehensive performance.

关 键 词：锂离子电池 健康状态(SOH)估计 恒压(CV)充电 灰狼优化(GWO)算法 支持向量机回归(SVR) 

分 类 号：TM912.9[电气工程—电力电子与电力传动]