锂离子电池极片表面缺陷检测方法研究进展  

Research progress on defect detection methods for electrodesurface of lithium-ion battery

作　　者：李博文 杨续来 葛肖尽 周帆 李渡阳 Li Bowen;Yang Xulai;Ge Xiaojin;Zhou Fan;Li Duyang(LIB Technology Center of Anhui Province,Hefei University,Hefei 230601,China;NETC Innovation Center,Hefei 230031,China)

机构地区：[1]合肥大学安徽省锂离子动力与储能电池产业共性技术研究中心,合肥230601 [2]国科能源技术创新中心,合肥230031

出　　处：《仪器仪表学报》2025年第1期125-146,共22页Chinese Journal of Scientific Instrument

基　　金：合肥市自然科学基金项目(2023042);安徽省重点研发计划项目(2023z04020004)资助。

摘　　要：极片作为锂离子电池的重要组件,在涂覆、辊压等环节中,表面容易产生划痕、露箔等缺陷,这些缺陷会严重影响电池的质量和使用寿命,从而使得电池极片表面缺陷检测和管控工序是锂离子电池生产过程中不可缺少的工艺环节。首先对锂离子电池极片的生产工艺进行介绍,并对生产过程中可能产生极片表面缺陷的原因和缺陷种类进行分析;然后阐述了用机器视觉代替人工对极片进行自动化检测的极片表面缺陷识别方法,主要介绍了传统机器视觉缺陷检测方法的原理以及优缺点,并深入分析了深度学习在极片表面缺陷检测领域中应用的原理和流程,同时对目标检测算法中的单、双阶段算法在锂离子电池极片表面缺陷检测中的应用进行重点分析与比较;最后对基于深度学习的机器视觉检测方法在锂离子电池极片表面缺陷检测中的未来发展方向进行展望,为该领域的研究人员提供更多参考。总的来说,极片表面缺陷检测技术的发展不仅依赖于工业相机等硬件设备的技术突破,更需要软件算法的不断优化和创新,软件和硬件的协同工作才能在保证检测精度的同时,提高检测效率和降低检测成本,进一步推动锂离子电池产业的高质量发展。Electrode is a crucial component of lithium-ion battery.In processes such as coating and rolling,the surface is prone to defects like scratches and foil exposure,these defects will severely affect the quality and service life of the battery.Therefore,the defect detection and control procedures for battery electrodes are indispensable process steps in the production of lithium-ion batteries.This article initially outlines the production process of lithium-battery electrode and analyzes the possible causes and types of surface defects of the electrodes during the production process.Then,it elaborates on the surface defect identification method for electrodes by using machine vision to replace manual labor for automated detection,it mainly introduces the principles,advantages,and disadvantages of the traditional machine vision defect detection method.Subsequently,it deeply analyzes the principles and procedures of the application of deep learning in the field of electrode surface defect detection,with a particular emphasis on the analysis and comparison of the applications of one-stage and two-stage algorithms in the target detection algorithm in lithium-ion battery electrode defect detection.Finally,the future development direction of the machine vision detection method based on deep learning is predicted for lithium-ion battery electrode surface defect detection,which can give more references for researchers in this domain.Overall,the development of electrode surface defect detection technology not only depends on the technological breakthroughs of hardware equipment such as industrial cameras,but also requires the continuous optimization and innovation of software algorithms.The synergy of software and hardware can ensure the detection accuracy,improve the detection efficiency and reduce the detection cost,then promote the lithium-ion battery production industry to embark on a high-quality development.Electrodes are critical components of lithium-ion batteries,and during manufacturing processes like coating and r

关 键 词：锂离子电池 极片表面缺陷 机器视觉 图像处理 深度学习 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TH89[自动化与计算机技术—计算机科学与技术]