基于产线数据驱动建模的锂离子电池分容技术

Translated title of the contribution: Capacity grading of lithium-ion battery based on model driven by production line data

毛宇, 冯雪松, 张晓琨, 向勇

Research output: Journal PublicationsJournal Article (refereed)peer-review

Abstract

针对简化电池分容工序的需求,提出了一种基于产线数据驱动机器学习建模的锂离子电池分容容量精准预测技术。采用2万余条源自同一锂离子电池制造产线的数据,应用四种非线性机器学习算法,分别探索了锂离子电池分容前产线监测数据与其分容容量之间的相关性。所有回归模型中,CatBoost算法模型表现出最优的容量预测精度,其测试集预测结果的均方根误差仅占标准容量的0.32%。此外,经过模型特征参量贡献度统计分析发现,OCV2和卷芯质量是分容容量预测的关键影响参数。
Translated title of the contributionCapacity grading of lithium-ion battery based on model driven by production line data
Original languageChinese (Simplified)
Pages (from-to)312-316
Number of pages5
Journal电源技术
Volume24
Issue number2
Publication statusPublished - Feb 2024
Externally publishedYes

Keywords

  • 锂离子电池
  • 产线数据
  • 分容
  • 数据驱动
  • CatBoost
  • lithium-ion batteries
  • production line data
  • capacity grading
  • data driving

Fingerprint

Dive into the research topics of 'Capacity grading of lithium-ion battery based on model driven by production line data'. Together they form a unique fingerprint.

Cite this