考虑历史退化信息融合的电池健康状态估计研究

打开文本图片集

主题词：锂离子电池 健康状态 特征提取

中图分类号：U46；TP18；TN303.1 文献标志码：A DOI：10.19620/j.cnki.1000-3703.20250111

ResearchonBatteryStateofHealth Estimationwith Historical DegradationInformation Fusion

ZhouDinghual，ZuoPeiwen²，ZhuZhongwen’，QiuXin'，MaQilong1 （1.School of Automotiveand Transportation Engineering，Hefei UniversityofTechnology，Hefei23ooo9;2.China AutomotiveInformation Technology（Tianjin） Co.，Ltd.， Tianjin，3O0000）

【Abstract】Inorder toaccurately estimate the State of Health （SOH）of lithium-ion bateries，this paper proposes an advanced SOHestimation methodthatintegrates Strategic Optimization Algorithm （SOA）with Memory-Enhanced Long ShortTermMemory （MELSTM）neuralnetwork.Firstly，a Variational AutoEncoder（VAE）isutilizedtoprocessrawdata，reducing redundant informationandextracting healthindicators，therebyachievingapreciserepresentationof baterydegradation information.Subsequently，ahybridmodelcombiningSOAandMELSTMisproposedtoestimateSOHoflithium-ionbatteries. Finall，effectivenessofteproposedmethodisvalidatedusing2publicdatasetsforlitium-onbateryaging，amelyACLE andNASA.Experimentalresultsdemonstratethattheproposed method improves RMSE indicators byover30%compared with conventional LSTMalgorithm，ofering new insights and solutionsforaccurateSOHestimationof ithium-ionbattery.

Keywords：Lithium-ionbattery，StateofHealth（SOH），Featureextraction 【引用格式】周定华，左培文，朱仲文，等.考虑历史退化信息融合的电池健康状态估计研究[J].汽车技术，2025（6）：28-35. ZHOUDH，ZUOPW，ZHUZW，etal.ResearchonBateryStateofHealthEstimation withHistoricalDegradation InformationFusion[J].Automobile Technology，2025（6）：28-35.

1前言

锂离子电池因具有自放电率低、循环寿命长、能量密度高、功率密度高、放电平稳、工作温度范围宽、无记忆效应和环保等优势[-3]，广泛应用于电动汽车、大型储能系统、航空等领域。（剩余11933字）