Intelligent Computed Tomography-based Detection Method for Lithium Battery Mylar Film Damage
-
摘要:
近年来,随着锂电池行业的快速发展与技术创新,电池的安全性能检测变得愈发重要。Mylar膜作为锂电池组装的重要组成部分,能极大地提升锂电池的使用安全性。然而,针对Mylar膜的破损检测研究却鲜有开展。因此,本文创新性地提出一种基于CT扫描的锂电池Mylar膜破损智能检测方法。该方法借助CT扫描这一无损检测技术,精准获取锂电池内部信息;随后结合图像预处理技术与深度学习算法,构建智能检测模型,实现对缺陷电池的高效、精准检测。实验结果表明,该方法对Mylar膜破损缺陷具有高检出率和低误检率,具有较高的应用价值。
-
关键词:
- 锂电池Mylar膜 /
- 缺陷检测 /
- Retinex图像增强 /
- 图像分类
Abstract:With the rapid development and innovation of the lithium battery industry in recent years, battery safety performance testing has become increasingly important. As an essential component of lithium batteries, Mylar films can significantly improve the safety of lithium batteries. However, few studies have focused on damage detection in Mylar films. To address this issue, this study developed an innovative intelligent detection method for lithium battery Mylar film damage. This method utilizes computed tomography (CT) nondestructive testing technology to accurately obtain internal information on lithium batteries. Subsequently, by combining image-preprocessing techniques and deep learning algorithms, an intelligent detection model was constructed to efficiently and accurately detect defective batteries. Experimental results demonstrate that the proposed method achieves a high detection rate and low false-detection rate for Mylar film defects, highlighting its significant potential for practical applications.
-
表 1
Mylar 膜破损的智能检测算法Table 1 Intelligent detection method for
Mylar film damage算法 1 Mylar膜破损的智能检测算法 1: 数据采集与重建:利用X射线CT扫描系统对锂电池进行扫描,采集投影图像序列,并通过三维重建算法得到高精度三维断层图像。 2: 切片提取与预处理:从断层图像中提取平行于Mylar膜方向的大面切片图像。对所选切片进行图像预处理,包括数据量化、对比度增强等操作,以突出Mylar膜破损特征。 3: 深度学习分类:基于预处理后的切片图像,选择神经网络进行特征提取和分类。 表 2 MSR对比度增强算法
Table 2 MSR contrast enhancement algorithm
算法 2 MSR(Multi-Scale Retinex)对比度增强算法 输入:图像I(x,y),高斯函数的尺度参数σ1,σ2,⋯,σN。 输出:对比度增强后的图像Ienhence(x,y)。 1. 对数域变换:I′(x,y)=logI(x,y)。 2. Ln(x,y)=Gn(x,y)∗I′(x,y),(n=1,2,⋯,N)。 3. Rn(x,y)=I′(x,y)−Ln(x,y)。 4. R(x,y)=∑Nn=1wnRn(x,y),其中∑Nn=1wn=1。 5. Ienhence(x,y)=RESIZE(exp(R(x,y)))。 注:Ln(x,y)为光照分量的估计值,Gn(x,y)为不同尺度σn的高斯函数,Rn(x,y)为每个尺度下的反射率估计值。RESIZE(⋅)为线性拉伸算子,第5步表示对R(x,y)进行指数变换并线性拉伸至原始图像范围得到对比度增强后的图像Ienhence(x,y)。 表 3 锂电池
CT 扫描重建条件Table 3 CT scanning and reconstruction conditions for lithium batteries
参数名称 参数值 电压 150 kV 电流 400 μA 曝光时间 0.05 s 帧间合并数 6 扫描角度 360° 扫描角度间隔 0.2° 探测器单元尺寸 0.15 mm×0.15 mm 探测器单元个数 2048 ×2048 射线源到旋转中心的距离 100 mm 射线源到探测器的距离 545 mm 重建图像尺寸 27.5 μm×27.5 μm×27.5 μm 重建图像体素个数 1140 ×1448 ×1536 表 4 数据集信息
Table 4 Dataset information
项目 训练集 验证集 测试集 总计 图像总数量 760 380 380 1520 破损样本数量 12 6 6 24 过采样的破损样本数量 360 60 0 420 -
[1] 何阳, 周永涛, 胡彬, 等. 一种导热阻燃型锂电池麦拉膜、其制备方法及高安全锂电池: CN202410802021.9[P]. 2024-10-22. [2] BAK S M, SHADIKE Z, LIN R, et al. In situ/operando synchrotron-based X-ray techniques for lithium-ion battery research [J]. NPG Asia Materials, 2018, 10: 563-580. DOI: 10.1038/s41427-018-0056-z.
[3] 田君, 田崔钧, 王一拓, 等. 锂离子电池安全性测试与评价方法分析[J]. 储能科学与技术, 2018, 7(6): 1128-1134. DOI: 10.12028/j.issn.2095-4239.2018.0154. TIAN J, TIAN C J, WANG Y T, et al. Safety test and evaluation method of lithium ion battery[J]. Energy Storage Science and Technology, 2018, 7(6): 1128-1134. DOI: 10.12028/j.issn.2095-4239.2018.0154. (in Chinese).
[4] 刘峰, 魏思伟, 蒋治亿, 等. 锂电池用 Mylar膜、锂电池和电子产品: CN202321619548.5[P]. 2024-02-06. [5] 刘大同, 周建宝, 郭力萌, 等. 锂离子电池健康评估和寿命预测综述[J]. 仪器仪表学报, 2015, 36(1): 1-16. LIU D T, ZHOU J B, GUO L M, et al. Survey on lithium-ion battery health assessment and cycle life estimation[J]. Chinese Journal of Scientific Instrument, 2015, 36(1): 1-16. (in Chinese).
[6] 王卓, 许江华, 张志勇, 等. 一种锂电池 Mylar膜缺陷检测方法: CN202210200989.5[P]. 2022-07-01. [7] 易凯, 钟鹏, 谢晏武, 等. 一种锂电池包 Mylar膜检测方法及检测系统, CN202111425784.9[P]. 2022-03-04. [8] HEENAN T M M, TAN C, WADE A J, et al. Theoretical transmissions for X-ray computed tomography studies of lithium-ion battery cathodes[J]. Materials & Design, 2020, 191: 108585. DOI: 10.1016/j.matdes.2020.108585.
[9] ALFRAHEED M I. A review of measurement methods for lithium-based battery defect and degradation analysis[J]. International Journal of Modelling and Simulation, 2025: 1-19. DOI: 10.1080/02286203.2025.2479667.
[10] 葛春平, 李育林, 曹琪. X射线检测在叠片式锂离子电池生产中的应用[J]. 电池, 2014, 44(4): 232-234. DOI: 10.3969/j.issn.1001-1579.2014.04.014. GE C P, LI Y L, CAO Q. Application of X-ray detection on laminated Li-ion battery production[J]. Battery Bimonthly, 2014, 44(4): 232-234. DOI: 10.3969/j.issn.1001-1579.2014.04.014. (in Chinese).
[11] 葛春平, 李育林, 薛渭萍, 等. 叠片式锂离子电池在线快速CT检测技术[J]. 电池, 2024, 54(3): 390-394. DOI: 10.19535/j.1001-1579.2024.03.020. GE C P, LI Y L, XUE W P. On-line rapid CT detection technology for laminated Li-ion battery[J]. Battery Bimonthly, 2024, 54(3): 390-394. DOI: 10.19535/j.1001-1579.2024.03.020. (in Chinese).
[12] KRIZHEVSKY A, SUTKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Association for Computing Machinery, 2017, 60(6): 84-90. DOI: 10.1145/3065386.
[13] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[J]. ArXiv, 2014. DOI: 10.48550/arXiv.1409.1556.
[14] SZEGEDY C, LIU W, JIA Y Q, et al. Going deeper with convolutions[C]//2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015: 1-9. DOI: 10.1109/CVPR.2015.7298594.
[15] HE K M, ZHANG X Y, REN S, et al. Deep residual learning for image recognition[C]//2016 IEEEConference on Computer Vision and Pattern Recognition (CVPR), 2016: 770-778. DOI: 10.1109/CVPR.2016.90.
[16] HUANG G, LIU Z, MAATEN L V D, et al. Densely connected convolutional networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 2261-2269. DOI: 10.1109/CVPR.2017.243.
[17] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16×16 words: Transformers for image recognition at scale[J]. ArXiv, 2020. DOI: 10.48550/arXiv.2010.11929.
[18] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[C]//IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149. DOI: 10.1109/TPAMI.2016.2577031.
[19] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016: 779-788. DOI: 10.1109/CVPR.2016.91.
[20] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023: 7464-7475. DOI: 10.1109/CVPR52729.2023.00721.
[21] AI Y, YE T D. Surface defect detection algorithm for PCB based on improved YOLOv8[C]//2024 8th International Conference on Electrical, Mechanical and Computer Engineering (ICEMCE), 2024: 1421-1425. DOI: 10.1109/ICEMCE64157.2024.10862076.
[22] RONNEBERGER O, FISCHER P, BROX T. U-Net: Convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention, 2015: 234-241. DOI: 10.48550/arXiv.1505.04597.
[23] BADRINARAYANAN V, KENDALL A, CIPOLLA R. SegNet: A deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495. DOI: 10.1109/TPAMI.2016.2644615.
[24] ZHAO H S, SHI J P, QI X J, et al. Pyramid scene parsing network[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017: 6230-6239. DOI: 10.1109/CVPR.2017.660.
[25] LAND E H, MCCANN J J. Lightness and retinex theory[J]. Journal of the Optical Society of America, 1971, 61(1): 1-11. DOI: 10.1364/josa.61.000001.
[26] PETRO A B, SBERT C, MOREL J M. Multiscale retinex[J]. Image Process on Line, 2014: 71-88. DOI: 10.5201/ipol.2014.107.
[27] PAN S J, YANG Q. A survey on transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(10): 1345-1359. DOI: 10.1109/TKDE.2009.191.
[28] YOSINSKI J, CLUNE J, BENGIO Y, et al. How transferable are features in deep neural networks[J] Advances in Neural Information Processing Systems (NIPS), 2014: 27. DOI: 10.48550/arXiv.1411.1792.
[29] 郑远攀, 李广阳, 李晔. 深度学习在图像识别中的应用研究综述[J]. 计算机工程与应用, 2019, 55(12): 20-36. DOI: 10.3778/j.issn.1002-8331.1903-0031. ZHENG Y P, LI G Y, LI Y. Survey of application of deep learning in image recognition[J]. Computer Engineering and Applications, 2019, 55(12): 20-36. DOI: 10.3778/j.issn.1002-8331.1903-0031. (in Chinese).
[30] WANG G, LIN T H, CHENG P, et al. A general cone-beam reconstruction algorithm[J]. IEEE Transactions on Meical Imaging, 1993, 12(3): 486-496. DOI: 10.1109/42.241876.