Article(id=1279511873359954289, tenantId=1146029695717560320, journalId=1278651732997652489, issueId=1279511628118986881, articleNumber=null, orderNo=null, doi=10.12086/oee.2026.250285, pmid=null, cstr=32245.14.oee.2026.250285, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1758556800000, receivedDateStr=2025-09-23, revisedDate=1765123200000, revisedDateStr=2025-12-08, acceptedDate=1765209600000, acceptedDateStr=2025-12-09, onlineDate=1782989002789, onlineDateStr=2026-07-02, pubDate=1776960000000, pubDateStr=2026-04-24, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1782989002789, onlineIssueDateStr=2026-07-02, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1782989002789, creator=13701087609, updateTime=1782989002789, updator=13701087609, issue=Issue{id=1279511628118986881, tenantId=1146029695717560320, journalId=1278651732997652489, year='2026', volume='53', issue='4', pageStart='250244', pageEnd='250340', issueExtLink='null', onlineDate='null', pubDate='1776960000000', pubDateStr='2026-04-24', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1782988944320, creator='13701087609', updateTime=1782988944320, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext=null, issueFiles=null, downloadFileDto=null}, startPage=250285, endPage=, ext={EN=ArticleExt(id=1279511876010754418, articleId=1279511873359954289, tenantId=1146029695717560320, journalId=1278651732997652489, language=EN, title=Raman spectroscopic aging assessment of oil-paper insulation based on an Inception-Resnet network integrating dilated convolution, columnId=1279511634116841602, journalTitle=Opto-Electronic Engineering, columnName=Article, runingTitle=null, highlight=null, articleAbstract=
Objective

Power transformer safe operation depends critically on oil-paper insulation condition. Traditional insulation aging detection approaches possess significant drawbacks, including long testing cycles, destructive procedures, and insufficient precision. Raman spectroscopy offers a rapid, non-destructive alternative by capturing molecular vibration characteristics associated with aging byproducts. However, conventional machine learning algorithms exhibit low efficiency, high computational cost, and weak generalization capabilities when processing high-dimensional Raman spectral data. Advanced two-dimensional convolutional neural networks demand excessive computational resources through artificial dimensionality expansion. To address these limitations, an intelligent insulation aging assessment approach utilizing a one-dimensional convolutional neural network (1D-CNN) integrated with a Dilated Inception-ResNet module is developed. The goal is to achieve accurate, rapid, and robust aging stage classification by automatically extracting multi-scale spectral features while mitigating gradient vanishing issues commonly found in deep networks, balancing computational cost and diagnostic performance effectively.

Methods

Accelerated thermal aging experiments generated three hundred oil-paper insulation samples. Mineral oil and kraft paper mixtures underwent continuous heating at 120 degrees Celsius for up to 480 hours. Samples were collected every 24 hours. Gas chromatography measured furfural content to establish ground truth labels, categorizing the samples into four distinct aging stages: initial, mid-term, late, and final. A portable Raman spectrometer collected molecular vibration spectra from the prepared samples. The laser power was set at 300 milliwatts with an excitation wavelength of 784.711 nanometers and an integration time of 500 milliseconds. Raw Raman spectral data underwent a rigorous serial preprocessing pipeline. Savitzky-Golay smoothing eliminated high-frequency noise interference. Subsequently, the adaptive iteratively reweighted Penalized Least Squares algorithm corrected baseline drift caused by fluorescence background interference. This preprocessing generated high-quality spectral data inputs. An enhanced 1D-CNN architecture was constructed. The core innovation involved designing a Dilated Inception-ResNet module. The network initially utilized a standard one-dimensional convolutional layer and max-pooling layer for preliminary feature mapping and dimensionality reduction. Two cascaded Dilated Inception-ResNet modules followed. Each module incorporated four parallel feature processing branches. The first branch utilized point convolutions for channel dimension linear transformations. The second and third branches applied a compress-expand strategy, using initial 1×1 convolutions followed by 1×3 and 1×5 one-dimensional dilated convolutions, respectively, to capture medium and long-range temporal dependencies without increasing parameter count. The fourth branch utilized max-pooling for significant feature retention. To prevent network degradation and gradient vanishing, adaptive residual connections linked the inputs and outputs of these modules, utilizing a 1×1 convolution for dimension matching when necessary. The network concluded with global flattening and fully connected layers for classification. Training utilized the AdamW optimizer and cross-entropy loss function.

Results and Discussions

Repeated random sampling validation experiments evaluated baseline model performance. The standard 1D-CNN achieved an average accuracy of 89.83% and a recall of 89.26%, outperforming traditional support vector machine and K-nearest neighbor classifiers. This demonstrated the inherent advantage of deep learning in automatically extracting representations from complex, high-dimensional spectral data without relying on manual feature engineering. Ablation studies verified the efficacy of the proposed Dilated Inception-ResNet architecture. The enhanced model achieved a maximum test set classification accuracy of 96.67%. This represented a significant absolute accuracy improvement of 6.67% over the original 1D-CNN and 3.34% over a standard Inception-1DCNN model without dilated convolutions or residual connections. The loss function curve demonstrated rapid and stable convergence within twenty epochs, confirming that the adaptive residual connections successfully facilitated smooth gradient backpropagation and eliminated gradient vanishing problems. Computational complexity analysis revealed that while parameters and floating-point operations increased moderately, the single-sample inference time remained exceptionally low at 0.1424 milliseconds, fully satisfying real-time monitoring requirements. Further extensive testing assessed model robustness and generalization capability under suboptimal data conditions. Three dataset configurations with varying total sample sizes and class distributions evaluated performance across different train-test split ratios. For a highly imbalanced dataset containing 230 samples, the proposed model maintained an average accuracy exceeding 92.5% across all split ratios. For a constrained small dataset containing only 170 samples, the average accuracy remained robust above 91.7%. These consistent performance metrics across varied data scenarios proved the multi-scale feature extraction mechanism successfully learned intrinsic physicochemical aging features rather than relying on statistical class distributions. The network architecture effectively prevented over-attention to majority classes and ensured reliable recognition of minority class samples representing critical severe aging stages.

Conclusions

The proposed Dilated Inception-ResNet 1D-CNN model provides a superior, non-destructive, and rapid diagnostic solution for oil-paper insulation aging assessment. Serial preprocessing techniques combining Savitzky-Golay smoothing and adaptive iteratively reweighted Penalized Least Squares algorithms significantly enhance Raman spectral data quality. The integration of multi-branch parallel dilated convolutions expands receptive fields for multi-scale feature extraction without escalating computational costs, while adaptive residual connections ensure stable deep network training. The model demonstrates exceptional classification accuracy, stability, and robustness, even when processing small or heavily imbalanced datasets. This intelligent diagnostic framework offers reliable technical support for transformer condition monitoring, predictive maintenance scheduling, and power system reliability assurance.

, authors=Sipeng Li, Fugen Song*, Tao Jin, authorsList=Sipeng Li, Fugen Song, Tao Jin, authorCompany=null, correspAuthors=Fugen Song, authorNote=null, correspAuthorsNote=
, copyrightStatement=Copyright © 2026 Opto-Electronic Engineering. All rights reserved., copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1279511923804848545, articleId=1279511873359954289, tenantId=1146029695717560320, journalId=1278651732997652489, language=CN, title=基于Inception-ResNet网络融合空洞卷积的油纸绝缘拉曼光谱老化评估, columnId=1279511637858160772, journalTitle=光电工程, columnName=科研论文, runingTitle=null, highlight=null, articleAbstract=

为解决传统油纸绝缘老化检测方法周期长、破坏性及精度不足,以及传统机器学习模型处理高维光谱数据效率低、泛化能力弱的问题,本文提出一种融合空洞卷积Inception-ResNet模块的改进型一维卷积神经网络 (1D-CNN)用于油纸绝缘老化状态的智能评估。通过热老化实验制备了300个不同老化阶段的油纸绝缘样本,并利用拉曼光谱仪采集其分子振动特征。采用S-G平滑与airPLS算法对光谱数据进行预处理。所提模型通过多分支并行空洞卷积提取多尺度特征,并结合自适应残差连接以缓解梯度消失。结果表明,该模型在测试集上的分类准确率达到96.67%,显著优于原始1D-CNN (90%)和Inception-1DCNN (93.33%)。在不平衡和小样本数据条件下,模型依然表现出优异的鲁棒性和泛化能力。

, authors=李思朋, 宋福根*, 金涛, authorsList=李思朋, 宋福根, 金涛, authorCompany=null, correspAuthors=宋福根, authorNote=

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

, correspAuthorsNote=
宋福根,
, copyrightStatement=版权所有©《光电工程》编辑部 2026, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=ZnEkJEtnzM+05euCA5UqOg==, magXml=8sZhPcxFUD0fW7L02VSFuw==, pdfUrl=null, pdf=gH5i/Ki4JHpjUnu2MTc0sw==, pdfFileSize=5022808, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=9IdW/Q22w9wUVMXt7SZdYw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=JUX1kQwaYUR+Au0f+3ElbQ==, mapNumber=null, fund=null)}, authors=[Author(id=1280951054955168037, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=820483429@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1280951055051637031, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951054955168037, language=EN, stringName=Sipeng Li, firstName=Sipeng, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055148106025, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951054955168037, language=CN, stringName=李思朋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=福州大学电气工程与自动化学院,福建 福州 350108, bio={"img":"5/vzOOKKowqyPiBYDiy89w==","content":"

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

"}, bioImg=5/vzOOKKowqyPiBYDiy89w==, bioContent=

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])]), Author(id=1280951055223603499, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=sfgalong@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1280951055336849709, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055223603499, language=EN, stringName=Fugen Song, firstName=Fugen, middleName=null, lastName=Song, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055454290222, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055223603499, language=CN, stringName=宋福根, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=福州大学电气工程与自动化学院,福建 福州 350108, bio={"img":"5g1b6Kmh/++ZGUCDo0Gr4w==","content":"

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

"}, bioImg=5g1b6Kmh/++ZGUCDo0Gr4w==, bioContent=

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])]), Author(id=1280951055525593392, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1280951055630450995, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055525593392, language=EN, stringName=Tao Jin, firstName=Tao, middleName=null, lastName=Jin, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055768863028, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055525593392, language=CN, stringName=金涛, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=福州大学电气工程与自动化学院,福建 福州 350108, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])])], keywords=[Keyword(id=1280951055919857974, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, orderNo=1, keyword=Oil-paper insulation), Keyword(id=1280951056024715575, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, orderNo=2, keyword=raman spectroscopy), Keyword(id=1280951056116990265, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, orderNo=3, keyword=Inception-ResNet network), Keyword(id=1280951056200876346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, orderNo=4, keyword=dilated convolution), Keyword(id=1280951056301539643, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, orderNo=5, keyword=aging assessment), Keyword(id=1280951056377037117, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, orderNo=1, keyword=油纸绝缘), Keyword(id=1280951056481894718, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, orderNo=2, keyword=拉曼光谱), Keyword(id=1280951056565780800, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, orderNo=3, keyword=Inception-ResNet网络), Keyword(id=1280951056716775744, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, orderNo=4, keyword=空洞卷积), Keyword(id=1280951056792273217, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, orderNo=5, keyword=老化评估)], refs=[Reference(id=1280951060613284213, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=梁栋, 朱建华, 张翠, 等. 变压器状态评估及故障诊断研究综述[J]. 变压器, 2024, 61 (2): 35−43., articleTitle=null, refAbstract=null), Reference(id=1280951060680393078, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Liang D, Zhu J H, Zhang C, et al. Review of transformer condition assessment and fault diagnosis[J]. Transformer, 2024, 61 (2): 35−43., articleTitle=null, refAbstract=null), Reference(id=1280951060760084855, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=高浩, 刘庆珍, 蔡金锭. 基于去极化电流Prony拟合的油纸绝缘德拜参数辨识方法[J]. 高压电器, 2020, 56 (11): 210−218., articleTitle=null, refAbstract=null), Reference(id=1280951060835582328, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Gao H, Liu Q Z, Cai J D. Debye parameter identification method of oil-paper insulation based on depolarization current Prony algorithm fitting[J]. High Voltage Appar, 2020, 56 (11): 210−218., articleTitle=null, refAbstract=null), Reference(id=1280951060986577273, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=陈啸轩, 邹阳, 翁祖辰, 等. 基于IKNN和LOF的变压器回复电压数据清洗方法研究[J]. 电子测量与仪器学报, 2024, 38 (2): 92−100., articleTitle=null, refAbstract=null), Reference(id=1280951061074657658, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen X X, Zou Y, Weng Z C, et al. Recovery voltage data cleaning method for transformer based on IKNN and LOF[J]. J Electron Meas Instrum, 2024, 38 (2): 92−100., articleTitle=null, refAbstract=null), Reference(id=1280951061162738043, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=陈钰林, 许再尧, 莫元雄, 等. 基于绝缘纸聚合度预测的油浸式变压器老化评估方法研究[J]. 红水河, 2023, 42 (5): 107−111., articleTitle=null, refAbstract=null), Reference(id=1280951061250818428, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen Y L, Xu Z Y, Mo Y X, et al. Aging evaluating method for oil immersed transformer based on insulating paper polymerization degree prediction[J]. Hongshui River, 2023, 42 (5): 107−111., articleTitle=null, refAbstract=null), Reference(id=1280951061338898813, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=李杰, 周渠, 贾路芬, 等. 红外、拉曼光谱的变压器油中糠醛检测方法对比研究[J]. 光谱学与光谱分析, 2024, 44 (1): 125−133., articleTitle=null, refAbstract=null), Reference(id=1280951061418590590, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Li J, Zhou Q, Jia L F, et al. Comparative study on detection methods of furfural in transformer oil based on IR and Raman spectroscopy[J]. Spectrosc Spectral Anal, 2024, 44 (1): 125−133., articleTitle=null, refAbstract=null), Reference(id=1280951061485699455, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=王加安, 刘立人, 李延, 等. 基于拉曼高光谱成像技术检测面粉中的偶氮甲酰胺[J]. 电子测量技术, 2022, 45 (14): 97−102., articleTitle=null, refAbstract=null), Reference(id=1280951061548614016, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang J A, Liu L R, Li Y, et al. Detection of azoformamide in flour based on Raman hyperspectral imaging[J]. Electron Meas Technol, 2022, 45 (14): 97−102., articleTitle=null, refAbstract=null), Reference(id=1280951061649277313, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=叶轲夫, 谢敏捷, 陈兴祺, 等. 拉曼光谱技术在环境微纳塑料检测中的应用与挑战[J]. 化学进展, 2025, 37 (1): 2−15., articleTitle=null, refAbstract=null), Reference(id=1280951061741552002, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Ye K F, Xie M J, Chen X Q, et al. Raman spectroscopy in the detection of environmental micro-and nanoplastics: applications and challenges[J]. Prog Chem, 2025, 37 (1): 2−15., articleTitle=null, refAbstract=null), Reference(id=1280951061838020995, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo N Z, Niu S J, Geng Y, et al. Non-destructive quantification of low colchicine concentrations in commercially available tablets using transmission Raman spectroscopy with partial least squares[J]. Int J Pharm: X, 2025, 9: 100321., articleTitle=null, refAbstract=null), Reference(id=1280951061934489988, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=杨定坤. 油纸绝缘老化拉曼光谱多尺度特征提取及集成增强神经网络诊断研究[D]. 重庆: 重庆大学, 2021. https://doi.org/10.27670/d.cnki.gcqdu.2021.000468., articleTitle=null, refAbstract=null), Reference(id=1280951062030958981, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang D K. Study on multi scale feature extraction and integrated enhanced neural network diagnosis of oil-paper insulation aging by Raman spectroscopy[D]. Chongqing: Chongqing University, 2021. https://doi.org/10.27670/d.cnki.gcqdu.2021.000468., articleTitle=null, refAbstract=null), Reference(id=1280951062127427974, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=陈新岗, 张文轩, 范益杰, 等. 基于局部线性嵌入的油纸绝缘拉曼光谱老化状态判别[J]. 激光与光电子学进展, 2025, 62 (3): 0330003., articleTitle=null, refAbstract=null), Reference(id=1280951062207119751, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen X G, Zhang W X, Fan Y J, et al. Determination of aging state of oil-paper insulation Raman spectrum based on local linear embedding[J]. Laser Optoelectron Prog, 2025, 62 (3): 0330003., articleTitle=null, refAbstract=null), Reference(id=1280951062291005832, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Fang J, Lin X, Zhou F X, et al. Site assessment of transformer state based on individual Raman spectrum equipment[J]. J Phys Conf Ser, 2023, 2584 (1): 012072., articleTitle=null, refAbstract=null), Reference(id=1280951062366503305, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=陈新岗, 范益杰, 马志鹏, 等. 基于集成增强KNN的油纸绝缘原始拉曼光谱老化状态判别[J]. 激光与光电子学进展, 2023, 60 (21): 2130002., articleTitle=null, refAbstract=null), Reference(id=1280951062479749514, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Chen X G, Fan Y J, Ma Z P, et al. Aging state discrimination of oil-paper insulation using raman spectroscopy and integrated enhanced KNN[J]. Laser Optoelectron Prog, 2023, 60 (21): 2130002., articleTitle=null, refAbstract=null), Reference(id=1280951062559441291, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=汪帮瑞, 薛建侠, 左小玉, 等. 基于拉曼光谱评估变压器油-屏障式绝缘老化状态[J]. 电气工程, 2023, 11 (2): 64−73., articleTitle=null, refAbstract=null), Reference(id=1280951062618161548, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang B R, Xue J X, Zuo X Y, et al. Evaluation of transformer oil-barrier based on Raman spectroscopy insulation aging state[J]. J Electr Eng, 2023, 11 (2): 64−73., articleTitle=null, refAbstract=null), Reference(id=1280951062714630541, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=郝明, 白鹤, 徐婷婷. 融合ResNeSt和多尺度特征融合的遥感影像道路提取[J]. 光电工程, 2025, 52 (1): 240236., articleTitle=null, refAbstract=null), Reference(id=1280951062785933710, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=Hao M, Bai H, Xu T T. Remote sensing image road extraction by integrating ResNeSt and multi-scale feature fusion[J]. Opto-Electron Eng, 2025, 52 (1): 240236., articleTitle=null, refAbstract=null), Reference(id=1280951062878208399, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=张晨晨, 王帅, 王文一, 等. 针对人脸识别卷积神经网络的局部背景区域对抗攻击[J]. 光电工程, 2023, 50 (1): 220266., articleTitle=null, refAbstract=null), Reference(id=1280951062974677392, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhang C C, Wang S, Wang W Y, et al. Adversarial background attacks in a limited area for CNN based face recognition[J]. Opto-Electron Eng, 2023, 50 (1): 220266., articleTitle=null, refAbstract=null), Reference(id=1280951063129866641, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Song R M, Chen W G, Yang D K, et al. Aging assessment of oil-paper insulation based on visional recognition of the dimensional expanded Raman spectra[J]. IEEE Trans Instrum Meas, 2021, 70: 6007110., articleTitle=null, refAbstract=null), Reference(id=1280951063247307154, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=钟明杉, 李兆飞, 张奕杰, 等. 基于一维卷积神经网络的天然气管道泄漏检测模型[J]. 国外电子测量技术, 2023, 42 (5): 62−68., articleTitle=null, refAbstract=null), Reference(id=1280951063364747667, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhong M S, Li Z F, Zhang Y J, et al. Natural gas pipeline leakage detection model based on 1-dimensional convolutional neural network[J]. For Electron Meas Technol, 2023, 42 (5): 62−68., articleTitle=null, refAbstract=null), Reference(id=1280951063469605268, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=伍济钢, 文港, 杨康. 改进一维卷积神经网络的航空发动机故障诊断方法[J]. 电子测量与仪器学报, 2023, 37 (3): 179−186., articleTitle=null, refAbstract=null), Reference(id=1280951063557685653, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=Wu J G, Wen G, Yang K. Improved one-dimensional convolutional neural network for aero-engine fault diagnosis[J]. J Electron Meas Instrum, 2023, 37 (3): 179−186., articleTitle=null, refAbstract=null), Reference(id=1280951063649960342, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=Xiong S Y, Wang C X, Zhu C S, et al. Dual detection of urea and glucose in sweat using a portable microfluidic SERS sensor with silver nano-tripods and 1D-CNN model analysis[J]. ACS Appl Mater Interfaces, 2024, 16 (48): 65918−65926., articleTitle=null, refAbstract=null), Reference(id=1280951063746429335, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=谢宇浩, 董前民, 金尚忠, 等. 基于深度神经网络的危险化学品拉曼光谱识别[J]. 激光与光电子学进展, 2025, 62 (5): 0530002., articleTitle=null, refAbstract=null), Reference(id=1280951063859675544, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=Xie Y H, Dong Q M, Jin S Z, et al. Raman spectroscopic identification of hazardous chemicals based on a deep neural network[J]. Laser Optoelectron Prog, 2025, 62 (5): 0530002., articleTitle=null, refAbstract=null), Reference(id=1280951063951950233, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=Xiong C C, Zhong Q S, Yan D H, et al. Multi-branch attention Raman network and surface-enhanced Raman spectroscopy for the classification of neurological disorders[J]. Biomed Opt Express, 2024, 15 (6): 3523−3540., articleTitle=null, refAbstract=null), Reference(id=1280951064027447706, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=Georgiev D, Fernández-Galiana Á, Pedersen S V, et al. Hyperspectral unmixing for Raman spectroscopy via physics-constrained autoencoders[J]. Proc Natl Acad Sci USA, 2024, 121 (45): e2407439121., articleTitle=null, refAbstract=null), Reference(id=1280951064107139483, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=涂潮, 刘万军, 赵琳琳, 等. 有限训练样本下的多尺度空洞密集网络高光谱影像分类[J]. 仪器仪表学报, 2024, 45 (4): 206−216., articleTitle=null, refAbstract=null), Reference(id=1280951064207802780, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=39, authorNames=null, journalName=null, refType=null, unstructuredReference=Tu C, Liu W J, Zhao L L, et al. Multiscale dilated dense network for hyperspectral image classification with limited training samples[J]. Chin J Sci Instrum, 2024, 45 (4): 206−216., articleTitle=null, refAbstract=null), Reference(id=1280951064279105949, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=40, authorNames=null, journalName=null, refType=null, unstructuredReference=李欢欢, 王思裕, 余武锟, 等. 绝缘纸在矿物油与菜籽油基天然酯中的加速热老化特性分析[J]. 绝缘材料, 2024, 57 (4): 44−48., articleTitle=null, refAbstract=null), Reference(id=1280951064346214814, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=41, authorNames=null, journalName=null, refType=null, unstructuredReference=Li H H, Wang S Y, Yu W K, et al. Comparative analysis on accelerated thermal ageing characteristics of insulating paper in mineral oil and rapeseed oil based natural ester[J]. Insul Mater, 2024, 57 (4): 44−48., articleTitle=null, refAbstract=null), Reference(id=1280951064434295199, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=42, authorNames=null, journalName=null, refType=null, unstructuredReference=郑成霞. airPLS算法去除拉曼光谱背景噪声的有效性研究[J]. 电子元器件与信息技术, 2021, 5 (2): 195−196., articleTitle=null, refAbstract=null), Reference(id=1280951064530764192, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=43, authorNames=null, journalName=null, refType=null, unstructuredReference=Zheng C X. Research on the effectiveness of airPLS algorithm in removing background noise from Raman spectra[J]. Electron Compon Inf Technol, 2021, 5 (2): 195−196, articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])], figs=[ArticleFig(id=1280951057018765635, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.1, caption=1D convolution operation, figureFileSmall=RmqqmOnyTwRy7qCVBqMaBw==, figureFileBig=9CUsEHopVYWIuAvlKQsr+Q==, tableContent=null), ArticleFig(id=1280951057106846021, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图1, caption=一维卷积操作, figureFileSmall=RmqqmOnyTwRy7qCVBqMaBw==, figureFileBig=9CUsEHopVYWIuAvlKQsr+Q==, tableContent=null), ArticleFig(id=1280951057228480838, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.2, caption=1D convolutional neural network (1D-CNN) architecture, figureFileSmall=PZJPSdiHLleJI+rSL3ZTRw==, figureFileBig=FUpFU3Uw1I8sRXFxKnuM6A==, tableContent=null), ArticleFig(id=1280951057308172616, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图2, caption=一维卷积神经网络结构图, figureFileSmall=PZJPSdiHLleJI+rSL3ZTRw==, figureFileBig=FUpFU3Uw1I8sRXFxKnuM6A==, tableContent=null), ArticleFig(id=1280951057400447305, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.3, caption=Schematic diagram of dilated convolution with different dilation rates, figureFileSmall=cuqH7XY2hzTfEbw3DA6kiw==, figureFileBig=uvwOSvV6rf+mSLjPLEKJew==, tableContent=null), ArticleFig(id=1280951057484333386, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图3, caption=不同膨胀率空洞卷积示意图, figureFileSmall=cuqH7XY2hzTfEbw3DA6kiw==, figureFileBig=uvwOSvV6rf+mSLjPLEKJew==, tableContent=null), ArticleFig(id=1280951057555636556, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.4, caption=Architecture of the dilated Inception-ResNet module, figureFileSmall=2sdmpG2lfSy34CMWQFjW5A==, figureFileBig=MNvFawzOS8LySVAz+M28jg==, tableContent=null), ArticleFig(id=1280951057631134029, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图4, caption=空洞Inception-ResNet模块结构图, figureFileSmall=2sdmpG2lfSy34CMWQFjW5A==, figureFileBig=MNvFawzOS8LySVAz+M28jg==, tableContent=null), ArticleFig(id=1280951057723408718, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.5, caption=Architecture of the 1D-CNN model based on the dilated convolution Inception-ResNet module, figureFileSmall=0Ub3BMdOz4wukaIRMTAPkQ==, figureFileBig=2kQMkmR4Jn8dvDYEAbZnSg==, tableContent=null), ArticleFig(id=1280951057798906191, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图5, caption=基于空洞卷积Inception-ResNet模块的1D-CNN模型架构, figureFileSmall=0Ub3BMdOz4wukaIRMTAPkQ==, figureFileBig=2kQMkmR4Jn8dvDYEAbZnSg==, tableContent=null), ArticleFig(id=1280951057874403664, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.6, caption=Photograph of aged oil samples, figureFileSmall=CIgsKOsznnshLEmvahmfXQ==, figureFileBig=J3HiWagQjJPHvwDNV1UDQw==, tableContent=null), ArticleFig(id=1280951057949901138, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图6, caption=老化油样实物图, figureFileSmall=CIgsKOsznnshLEmvahmfXQ==, figureFileBig=J3HiWagQjJPHvwDNV1UDQw==, tableContent=null), ArticleFig(id=1280951058017010003, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.7, caption=Schematic diagram of Raman spectroscopy detection, figureFileSmall=ITwBa0PV02pXMpHjsNKjAw==, figureFileBig=WliEQ1WGGdE5EqWTJ5eEGg==, tableContent=null), ArticleFig(id=1280951058096701780, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图7, caption=拉曼光谱检测示意图, figureFileSmall=ITwBa0PV02pXMpHjsNKjAw==, figureFileBig=WliEQ1WGGdE5EqWTJ5eEGg==, tableContent=null), ArticleFig(id=1280951058180587862, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.8, caption=Original Raman spectra of samples, figureFileSmall=SF6LeaUABCBPWOOm8QSUbg==, figureFileBig=E4hts2a8O46wsYKEjgd4mQ==, tableContent=null), ArticleFig(id=1280951058306416983, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图8, caption=样本的原始拉曼光谱图, figureFileSmall=SF6LeaUABCBPWOOm8QSUbg==, figureFileBig=E4hts2a8O46wsYKEjgd4mQ==, tableContent=null), ArticleFig(id=1280951058402885977, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.9, caption=Preprocessed Raman spectra, figureFileSmall=9rimSYWNQpXk1yj07jDM8A==, figureFileBig=Y0t53YLVkm179ycF5e2/NQ==, tableContent=null), ArticleFig(id=1280951058516132185, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图9, caption=预处理后的拉曼光谱图, figureFileSmall=9rimSYWNQpXk1yj07jDM8A==, figureFileBig=Y0t53YLVkm179ycF5e2/NQ==, tableContent=null), ArticleFig(id=1280951058595823962, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.10, caption=Overall workflow diagram of the model, figureFileSmall=QZTh+m7Tph5GkYb60IsOaQ==, figureFileBig=w7zlhJbX9m26/ues84yDcA==, tableContent=null), ArticleFig(id=1280951058679710044, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图10, caption=模型整体工作流程图, figureFileSmall=QZTh+m7Tph5GkYb60IsOaQ==, figureFileBig=w7zlhJbX9m26/ues84yDcA==, tableContent=null), ArticleFig(id=1280951058751013213, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.11, caption=Accuracy variation across multiple experiments, figureFileSmall=IswRdqkzl384zfDgzuP+XQ==, figureFileBig=IjrJQQb3Kz/4uhH5FbT3bA==, tableContent=null), ArticleFig(id=1280951058843287902, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图11, caption=多次实验准确率变化图, figureFileSmall=IswRdqkzl384zfDgzuP+XQ==, figureFileBig=IjrJQQb3Kz/4uhH5FbT3bA==, tableContent=null), ArticleFig(id=1280951058935562592, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.12, caption=Training loss, figureFileSmall=idd7k2JqEQ2bw1FtKn6umw==, figureFileBig=0R0YuWxPFWXLeA2A1DML4A==, tableContent=null), ArticleFig(id=1280951059011060065, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图12, caption=训练损失, figureFileSmall=idd7k2JqEQ2bw1FtKn6umw==, figureFileBig=0R0YuWxPFWXLeA2A1DML4A==, tableContent=null), ArticleFig(id=1280951059094946146, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Fig.13, caption=Confusion matrix for aging state discrimination. (a) Original 1D-CNN; (b) Inception-1DCNN; (c) The propsed model, figureFileSmall=aO8ozPFMDSi9JnbrhWiYsQ==, figureFileBig=em2pi84LpXG2eQ7nGi4E4g==, tableContent=null), ArticleFig(id=1280951059170443620, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=图13, caption=老化状态判别混淆矩阵。 (a) 原始的1D-CNN;(b) Inception-1DCNN;(c) 本文模型, figureFileSmall=aO8ozPFMDSi9JnbrhWiYsQ==, figureFileBig=em2pi84LpXG2eQ7nGi4E4g==, tableContent=null), ArticleFig(id=1280951059266912613, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.1, caption=

Classification of aging stages

, figureFileSmall=null, figureFileBig=null, tableContent=
Heating duration/hFurfural content (×10−6)Aging stage
0−1200−1Initial
120−2401.0−1.7Mid-term
240−3601.7−2.0Late
360−480>2Final
), ArticleFig(id=1280951059388547431, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表1, caption=

老化阶段分类

, figureFileSmall=null, figureFileBig=null, tableContent=
Heating duration/hFurfural content (×10−6)Aging stage
0−1200−1Initial
120−2401.0−1.7Mid-term
240−3601.7−2.0Late
360−480>2Final
), ArticleFig(id=1280951059480822119, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.2, caption=

Branch parameters of the dilated Inception-ResNet module

, figureFileSmall=null, figureFileBig=null, tableContent=
NameKernel sizeStrideChannelsPaddingDilation rate
1x1 Conv branch1×113201
1x3 dilated Conv branch1×313222
1x5 dilated Conv branch1×511642
Pool branch5×11162-
), ArticleFig(id=1280951059547930984, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表2, caption=

空洞Inception-ResNet模块分支参数

, figureFileSmall=null, figureFileBig=null, tableContent=
NameKernel sizeStrideChannelsPaddingDilation rate
1x1 Conv branch1×113201
1x3 dilated Conv branch1×313222
1x5 dilated Conv branch1×511642
Pool branch5×11162-
), ArticleFig(id=1280951059636011370, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.3, caption=

Performance comparison of discrimination models (unit: %)

, figureFileSmall=null, figureFileBig=null, tableContent=
Diagnostic modelAverage accuracyAccuracy stdRecall
1D-CNN89.832.5489.26
KNN87.504.9285.85
SVM87.503.7986.43
), ArticleFig(id=1280951059711508843, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表3, caption=

判别模型效果对比(单位:%)

, figureFileSmall=null, figureFileBig=null, tableContent=
Diagnostic modelAverage accuracyAccuracy stdRecall
1D-CNN89.832.5489.26
KNN87.504.9285.85
SVM87.503.7986.43
), ArticleFig(id=1280951059778617708, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.4, caption=

Comparison of results for the three models/%

, figureFileSmall=null, figureFileBig=null, tableContent=
NO.ModelAccuracyRecallF1 score
1Original 1D-CNN88.3387.0786.57
Inception-1DCNN91.6790.6690.69
Proposed model93.3393.3393.25
2Original 1D-CNN90.0090.6290.04
Inception-1DCNN91.6793.4291.90
Proposed model96.6796.6796.74
3Original 1D-CNN93.3393.3292.71
Inception-1DCNN95.0095.5994.58
Proposed model100100100
4Original 1D-CNN90.0090.3290.06
Inception-1DCNN93.3392.9493.20
Proposed model96.6797.0696.77
5Original 1D-CNN86.6786.9882.72
Inception-1DCNN90.0091.1889.29
Proposed model93.3392.9886.39
), ArticleFig(id=1280951059854115182, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表4, caption=

三种模型结果对比/%

, figureFileSmall=null, figureFileBig=null, tableContent=
NO.ModelAccuracyRecallF1 score
1Original 1D-CNN88.3387.0786.57
Inception-1DCNN91.6790.6690.69
Proposed model93.3393.3393.25
2Original 1D-CNN90.0090.6290.04
Inception-1DCNN91.6793.4291.90
Proposed model96.6796.6796.74
3Original 1D-CNN93.3393.3292.71
Inception-1DCNN95.0095.5994.58
Proposed model100100100
4Original 1D-CNN90.0090.3290.06
Inception-1DCNN93.3392.9493.20
Proposed model96.6797.0696.77
5Original 1D-CNN86.6786.9882.72
Inception-1DCNN90.0091.1889.29
Proposed model93.3392.9886.39
), ArticleFig(id=1280951059933806959, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.5, caption=

Model complexity and efficiency comparison

, figureFileSmall=null, figureFileBig=null, tableContent=
ModelParameters/MFLOPs/GInference time per sample/msTotal training time/s
Original 1D-CNN0.130.760.01671.18
Inception-1DCNN0.209.070.08332.28
Proposed model0.2123.020.14244.14
), ArticleFig(id=1280951060021887344, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表5, caption=

模型复杂度与效率对比

, figureFileSmall=null, figureFileBig=null, tableContent=
ModelParameters/MFLOPs/GInference time per sample/msTotal training time/s
Original 1D-CNN0.130.760.01671.18
Inception-1DCNN0.209.070.08332.28
Proposed model0.2123.020.14244.14
), ArticleFig(id=1280951060114162033, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.6, caption=

Sample combinations

, figureFileSmall=null, figureFileBig=null, tableContent=
GroupInitialMid-termLateFinalTotal
A75757575300
B50757530230
C35456030170
), ArticleFig(id=1280951060219019634, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表6, caption=

样本组合

, figureFileSmall=null, figureFileBig=null, tableContent=
GroupInitialMid-termLateFinalTotal
A75757575300
B50757530230
C35456030170
), ArticleFig(id=1280951060323877235, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=EN, label=Tab.7, caption=

Model discrimination results on different datasets (unit: %)

, figureFileSmall=null, figureFileBig=null, tableContent=
Average accuracyAverage recallF1 scoreAverage accuracyAverage recallF1 score
Ratio 9∶1Ratio 8.5∶1.5
A95.3394.8894.9294.2293.8793.61
B93.0494.0693.1092.5793.0893.16
C92.9492.0091.7195.3893.3391.88
Average accuracyAverage recallF1 ScoreAverage accuracyAverage recallF1 score
Ratio 8∶2Ratio 7∶3
A95.3395.4395.4193.1193.0893.04
B92.6193.4291.1993.0492.5592.26
C91.7790.7690.7191.7790.7890.57
), ArticleFig(id=1280951060432929140, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, language=CN, label=表7, caption=

不同数据集中模型判别结果(单位:%)

, figureFileSmall=null, figureFileBig=null, tableContent=
Average accuracyAverage recallF1 scoreAverage accuracyAverage recallF1 score
Ratio 9∶1Ratio 8.5∶1.5
A95.3394.8894.9294.2293.8793.61
B93.0494.0693.1092.5793.0893.16
C92.9492.0091.7195.3893.3391.88
Average accuracyAverage recallF1 ScoreAverage accuracyAverage recallF1 score
Ratio 8∶2Ratio 7∶3
A95.3395.4395.4193.1193.0893.04
B92.6193.4291.1993.0492.5592.26
C91.7790.7690.7191.7790.7890.57
)], attaches=null, journal=Journal(id=1278641367198941188, delFlag=0, nameCn=光电工程, nameEn=Opto-Electronic Engineering, nameHistory1=null, nameHistory2=null, issn=1003-501X, eissn=2097-4019, cn=51-1346/O4, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=4Vimkd+qXLWNxtdpr9mFNw==, journalPrice=null, startedYear=null, abbrevIsoEn=Opto-Electronic Engineering, journalRemark=null, publicationField=null, createdTime=1782781457950, updatedTime=1784021784694, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=G, firstLetterEn=G, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=4Vimkd+qXLWNxtdpr9mFNw==, picEn=vAq9s20WLs1ODDfbWq+Gjg==, jcr=null, cjcr=null, exts=[JournalExt(id=1283843675721536154, language=CN, name=光电工程, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1784021784954, updatedTime=1784021784954, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.manuscripts.com.cn/gdgc, submissionEditorUrl=http://www.manuscripts.com.cn/gdgc, submissionReviewUrl=http://www.manuscripts.com.cn/gdgc, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1283843675771867803, language=EN, name=Opto-Electronic Engineering, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1784021784966, updatedTime=1784021784966, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.manuscripts.com.cn/gdgc, submissionEditorUrl=http://www.manuscripts.com.cn/gdgc, submissionReviewUrl=http://www.manuscripts.com.cn/gdgc, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1278651732997652489, websiteList=[Website(id=1278723867418018151, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1278651732997652489, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/oee/CN, language=CN, createTime=1782801127533, createBy=18614031015, updateTime=1782804494442, updateBy=18614031015, name=光电工程-中文, tplId=1146099689490845704, title=光电工程, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1278738091150128034, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=articleTextType, value=kx, createTime=1782804518735, updateTime=1782804518735, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091120767903, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=banner, value=null, createTime=1782804518728, updateTime=1782804518728, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091171099557, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=grayFlag, value=0, createTime=1782804518740, updateTime=1782804518740, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091108184990, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=logo, value=https://castjournals.cast.org.cn/joweb/oee/CN/file/pic?fileId=A1C6uwqtMazluiWkEpR0Mg==, createTime=1782804518725, updateTime=1782804518725, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091179488167, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=minRunFlag, value=0, createTime=1782804518742, updateTime=1782804518742, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091141739425, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/oee/CN/file/pic, createTime=1782804518733, updateTime=1782804518733, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091175293862, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=silenceFlag, value=0, createTime=1782804518741, updateTime=1782804518741, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091129156512, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1782804518730, updateTime=1782804518730, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091154322339, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=themeColor, value=null, createTime=1782804518736, updateTime=1782804518736, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738091162710948, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867418018151, code=themeStyle, value=null, createTime=1782804518738, updateTime=1782804518738, creator=18614031015, updator=18614031015)]), Website(id=1278723867522875769, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1278651732997652489, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/oee/EN, language=EN, createTime=1782801127558, createBy=18614031015, updateTime=1782804490442, updateBy=18614031015, name=光电工程-英文, tplId=1146101810881728533, title=Opto-Electronic Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1278738063660659607, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=articleTextType, value=kx, createTime=1782804512181, updateTime=1782804512181, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063635493780, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=banner, value=null, createTime=1782804512175, updateTime=1782804512175, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063924900762, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=grayFlag, value=0, createTime=1782804512244, updateTime=1782804512244, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063606133651, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=logo, value=https://castjournals.cast.org.cn/joweb/oee/EN/file/pic?fileId=A1C6uwqtMazluiWkEpR0Mg==, createTime=1782804512168, updateTime=1782804512168, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063937483676, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=minRunFlag, value=0, createTime=1782804512247, updateTime=1782804512247, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063652270998, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/oee/EN/file/pic, createTime=1782804512179, updateTime=1782804512179, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063933289371, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=silenceFlag, value=0, createTime=1782804512246, updateTime=1782804512246, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063639688085, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1782804512176, updateTime=1782804512176, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063664853912, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=themeColor, value=null, createTime=1782804512182, updateTime=1782804512182, creator=18614031015, updator=18614031015), WebsiteProps(id=1278738063916512153, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1278723867522875769, code=themeStyle, value=null, createTime=1782804512242, updateTime=1782804512242, creator=18614031015, updator=18614031015)])], journalTitle=光电工程, weixinUrl=null, journalUrl=https://www.oejournal.org/oee, iacademicId=null, status=1, seqNo=null, journalTitleEn=Opto-Electronic Engineering, journalPhotoCn=4Vimkd+qXLWNxtdpr9mFNw==, journalPhotoEn=vAq9s20WLs1ODDfbWq+Gjg==, journalFirstLetter=G, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/oee/CN/10.12086/oee.2026.250285, detailUrlEn=https://castjournals.cast.org.cn/joweb/oee/EN/10.12086/oee.2026.250285, pdfUrlCn=https://castjournals.cast.org.cn/joweb/oee/CN/PDF/10.12086/oee.2026.250285, pdfUrlEn=https://castjournals.cast.org.cn/joweb/oee/EN/PDF/10.12086/oee.2026.250285, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1776960000000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于Inception-ResNet网络融合空洞卷积的油纸绝缘拉曼光谱老化评估
收藏切换
PDF下载
李思朋 , 宋福根 * , 金涛
光电工程 | 科研论文 2026,53(4): 250285
收起
收藏切换
光电工程 |科研论文 2026 , 53 (4) : 250285
基于Inception-ResNet网络融合空洞卷积的油纸绝缘拉曼光谱老化评估
全屏
[Author(id=1280951054955168037, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=820483429@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1280951055051637031, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951054955168037, language=EN, stringName=Sipeng Li, firstName=Sipeng, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055148106025, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951054955168037, language=CN, stringName=李思朋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=福州大学电气工程与自动化学院,福建 福州 350108, bio={"img":"5/vzOOKKowqyPiBYDiy89w==","content":"

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

"}, bioImg=5/vzOOKKowqyPiBYDiy89w==, bioContent=

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])]), Author(id=1280951055223603499, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=sfgalong@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1280951055336849709, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055223603499, language=EN, stringName=Fugen Song, firstName=Fugen, middleName=null, lastName=Song, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055454290222, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055223603499, language=CN, stringName=宋福根, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=福州大学电气工程与自动化学院,福建 福州 350108, bio={"img":"5g1b6Kmh/++ZGUCDo0Gr4w==","content":"

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

"}, bioImg=5g1b6Kmh/++ZGUCDo0Gr4w==, bioContent=

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])]), Author(id=1280951055525593392, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1280951055630450995, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055525593392, language=EN, stringName=Tao Jin, firstName=Tao, middleName=null, lastName=Jin, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1280951055768863028, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, authorId=1280951055525593392, language=CN, stringName=金涛, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=福州大学电气工程与自动化学院,福建 福州 350108, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1280951054854504737, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, xref=null, ext=[AuthorCompanyExt(id=1280951054862893346, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China), AuthorCompanyExt(id=1280951054871281955, tenantId=1146029695717560320, journalId=1278651732997652489, articleId=1279511873359954289, companyId=1280951054854504737, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=福州大学电气工程与自动化学院,福建 福州 350108)])])]
李思朋 , 宋福根* , 金涛
作者信息
  • 福州大学电气工程与自动化学院,福建 福州 350108
通讯作者:
宋福根,
作者简介:

李思朋 (2000-),男,硕士研究生,主要研究方向为油纸绝缘变压器的老化评估。E-mail:

宋福根 (1982-),男,博士,副教授,主要研究方向为电力系统智能化故障诊断和新型电力系统分析。Email:

Raman spectroscopic aging assessment of oil-paper insulation based on an Inception-Resnet network integrating dilated convolution
Sipeng Li , Fugen Song* , Tao Jin
Affiliations
  • College of Electrical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108,China
出版时间: 2026-04-24 doi: 10.12086/oee.2026.250285
文章导航
收藏切换

为解决传统油纸绝缘老化检测方法周期长、破坏性及精度不足,以及传统机器学习模型处理高维光谱数据效率低、泛化能力弱的问题,本文提出一种融合空洞卷积Inception-ResNet模块的改进型一维卷积神经网络 (1D-CNN)用于油纸绝缘老化状态的智能评估。通过热老化实验制备了300个不同老化阶段的油纸绝缘样本,并利用拉曼光谱仪采集其分子振动特征。采用S-G平滑与airPLS算法对光谱数据进行预处理。所提模型通过多分支并行空洞卷积提取多尺度特征,并结合自适应残差连接以缓解梯度消失。结果表明,该模型在测试集上的分类准确率达到96.67%,显著优于原始1D-CNN (90%)和Inception-1DCNN (93.33%)。在不平衡和小样本数据条件下,模型依然表现出优异的鲁棒性和泛化能力。

油纸绝缘  /  拉曼光谱  /  Inception-ResNet网络  /  空洞卷积  /  老化评估
Objective

Power transformer safe operation depends critically on oil-paper insulation condition. Traditional insulation aging detection approaches possess significant drawbacks, including long testing cycles, destructive procedures, and insufficient precision. Raman spectroscopy offers a rapid, non-destructive alternative by capturing molecular vibration characteristics associated with aging byproducts. However, conventional machine learning algorithms exhibit low efficiency, high computational cost, and weak generalization capabilities when processing high-dimensional Raman spectral data. Advanced two-dimensional convolutional neural networks demand excessive computational resources through artificial dimensionality expansion. To address these limitations, an intelligent insulation aging assessment approach utilizing a one-dimensional convolutional neural network (1D-CNN) integrated with a Dilated Inception-ResNet module is developed. The goal is to achieve accurate, rapid, and robust aging stage classification by automatically extracting multi-scale spectral features while mitigating gradient vanishing issues commonly found in deep networks, balancing computational cost and diagnostic performance effectively.

Methods

Accelerated thermal aging experiments generated three hundred oil-paper insulation samples. Mineral oil and kraft paper mixtures underwent continuous heating at 120 degrees Celsius for up to 480 hours. Samples were collected every 24 hours. Gas chromatography measured furfural content to establish ground truth labels, categorizing the samples into four distinct aging stages: initial, mid-term, late, and final. A portable Raman spectrometer collected molecular vibration spectra from the prepared samples. The laser power was set at 300 milliwatts with an excitation wavelength of 784.711 nanometers and an integration time of 500 milliseconds. Raw Raman spectral data underwent a rigorous serial preprocessing pipeline. Savitzky-Golay smoothing eliminated high-frequency noise interference. Subsequently, the adaptive iteratively reweighted Penalized Least Squares algorithm corrected baseline drift caused by fluorescence background interference. This preprocessing generated high-quality spectral data inputs. An enhanced 1D-CNN architecture was constructed. The core innovation involved designing a Dilated Inception-ResNet module. The network initially utilized a standard one-dimensional convolutional layer and max-pooling layer for preliminary feature mapping and dimensionality reduction. Two cascaded Dilated Inception-ResNet modules followed. Each module incorporated four parallel feature processing branches. The first branch utilized point convolutions for channel dimension linear transformations. The second and third branches applied a compress-expand strategy, using initial 1×1 convolutions followed by 1×3 and 1×5 one-dimensional dilated convolutions, respectively, to capture medium and long-range temporal dependencies without increasing parameter count. The fourth branch utilized max-pooling for significant feature retention. To prevent network degradation and gradient vanishing, adaptive residual connections linked the inputs and outputs of these modules, utilizing a 1×1 convolution for dimension matching when necessary. The network concluded with global flattening and fully connected layers for classification. Training utilized the AdamW optimizer and cross-entropy loss function.

Results and Discussions

Repeated random sampling validation experiments evaluated baseline model performance. The standard 1D-CNN achieved an average accuracy of 89.83% and a recall of 89.26%, outperforming traditional support vector machine and K-nearest neighbor classifiers. This demonstrated the inherent advantage of deep learning in automatically extracting representations from complex, high-dimensional spectral data without relying on manual feature engineering. Ablation studies verified the efficacy of the proposed Dilated Inception-ResNet architecture. The enhanced model achieved a maximum test set classification accuracy of 96.67%. This represented a significant absolute accuracy improvement of 6.67% over the original 1D-CNN and 3.34% over a standard Inception-1DCNN model without dilated convolutions or residual connections. The loss function curve demonstrated rapid and stable convergence within twenty epochs, confirming that the adaptive residual connections successfully facilitated smooth gradient backpropagation and eliminated gradient vanishing problems. Computational complexity analysis revealed that while parameters and floating-point operations increased moderately, the single-sample inference time remained exceptionally low at 0.1424 milliseconds, fully satisfying real-time monitoring requirements. Further extensive testing assessed model robustness and generalization capability under suboptimal data conditions. Three dataset configurations with varying total sample sizes and class distributions evaluated performance across different train-test split ratios. For a highly imbalanced dataset containing 230 samples, the proposed model maintained an average accuracy exceeding 92.5% across all split ratios. For a constrained small dataset containing only 170 samples, the average accuracy remained robust above 91.7%. These consistent performance metrics across varied data scenarios proved the multi-scale feature extraction mechanism successfully learned intrinsic physicochemical aging features rather than relying on statistical class distributions. The network architecture effectively prevented over-attention to majority classes and ensured reliable recognition of minority class samples representing critical severe aging stages.

Conclusions

The proposed Dilated Inception-ResNet 1D-CNN model provides a superior, non-destructive, and rapid diagnostic solution for oil-paper insulation aging assessment. Serial preprocessing techniques combining Savitzky-Golay smoothing and adaptive iteratively reweighted Penalized Least Squares algorithms significantly enhance Raman spectral data quality. The integration of multi-branch parallel dilated convolutions expands receptive fields for multi-scale feature extraction without escalating computational costs, while adaptive residual connections ensure stable deep network training. The model demonstrates exceptional classification accuracy, stability, and robustness, even when processing small or heavily imbalanced datasets. This intelligent diagnostic framework offers reliable technical support for transformer condition monitoring, predictive maintenance scheduling, and power system reliability assurance.

Oil-paper insulation  /  raman spectroscopy  /  Inception-ResNet network  /  dilated convolution  /  aging assessment
李思朋, 宋福根, 金涛. 基于Inception-ResNet网络融合空洞卷积的油纸绝缘拉曼光谱老化评估. 光电工程, 2026 , 53 (4) : 250285 - . DOI: 10.12086/oee.2026.250285
Sipeng Li, Fugen Song, Tao Jin. Raman spectroscopic aging assessment of oil-paper insulation based on an Inception-Resnet network integrating dilated convolution[J]. Opto-Electronic Engineering, 2026 , 53 (4) : 250285 - . DOI: 10.12086/oee.2026.250285
在电力系统的运行中,变压器的绝缘性能是保障电网安全运行的关键因素。油纸绝缘作为变压器的主要绝缘形式,长期运行会使其内部结构老化,进而导致绝缘性能下降,甚至引发电气故障。因此,及时准确地评估变压器的老化程度,对于电力系统安全极为重要[1-3]。然而,现有的老化检测方法,如糠醛含量检测、绝缘纸聚合度分析以及对微量老化特征物 (如CO、CO2等)的检测等[4-6],普遍存在检测周期长、操作复杂、对设备有破坏性等不足,难以适应现代电力系统对油纸绝缘老化状态快速、精准监测的需求。
近年来,拉曼光谱技术因其能够通过检测材料的分子振动特征,反映其化学成分的变化,被广泛应用于化学材料、食品安全和生物医学等领域[7-8]。在变压器油纸绝缘的老化过程中,会生成一些老化特征物并溶解在绝缘油中,这些特征物的种类和含量会随着老化阶段的不同而发生变化,因此采用拉曼光谱对其进行检测,通过光谱信号可以清晰地反映出这些差异,从而评估老化程度。相比传统检测方法,拉曼光谱技术具有快速、灵敏、非破坏性等优点,可实现对油纸绝缘老化特征物的快速检测,且检测过程无需复杂的油气分离。此外,结合机器学习和高维数据处理技术,拉曼光谱还可进一步挖掘老化特征,为变压器绝缘状态的精准评估提供新的视角[9-10]
目前,基于拉曼光谱对油纸绝缘变压器老化评估等已经有一定的研究,文献[11]使用MDS提取了能够反映变压器状态的八维拉曼光谱特征,并采用随机森林法设计了基于拉曼光谱的变压器状态评估模型,尽管随机森林结合拉曼光谱特征在变压器状态评估中具有一定的优势,但其在模型复杂度、数据噪声敏感性、特征选择局限性以及样本不平衡问题等方面仍存在不足;文献[12]使用AA-KNN构建针对原始高维拉曼光谱样本的判别模型,并与KNN、RF、XGBoost等算法进行了对比,证明其有着更好的性能;文献[13]利用主成分分析法 (PCA)对原始拉曼光谱数据进行降维,利用支持向量机 (SVM)搭建判别模型,并使用网格法、粒子群算法、遗传算法对支持向量机进行优化。这些传统机器学习算法主要是通过对二维数据矩阵的处理,分析不同老化程度样品之间的邻近关系来构建模型,但这些模型在复杂数据处理能力、模型可扩展性、模型复杂度和泛化能力等方面存在劣势,且对于拉曼光谱这种高维特征数据的处理效率较低,过于依赖数据降维技术,因此具有一定的局限性。
卷积神经网络 (Convolutional neural network, CNN)通过对光谱信号中的空间和时序特征自动学习,能够利用卷积层和池化层等对高维光谱特征进行提取,展现出强大的分类识别能力[14-15]。文献[16]将一维拉曼光谱数据通过一种基于本征乘积算术的变换方法转换为二维图像数据,借助基于增维法的CNN实现了基于拉曼光谱的不同老化状态的油纸绝缘诊断,但维度的增加必然需要更大的算力,使得计算成本和硬件要求大幅提高。而相较于2D-CNN,一维卷积神经网络 (1D-CNN)的卷积操作是在一维序列上进行的,能够直接有效捕获序列数据中的局部特征和模式,使得计算成本大幅降低,对于本身就是一维的变压器油纸绝缘原始拉曼光谱的识别有着明显的优势[17-19]。近年来,为了进一步提升模型性能,一些研究开始探索将更先进的深度学习结构运用在拉曼光谱的识别中。例如,基于注意力机制 (attention mechanism)的模型被用于自动学习光谱中的关键特征频移,从而增强了模型的判别能力[20-21],然而,这些模型通常结构更复杂,对训练数据量的要求也更高,且对于拉曼光谱中普遍存在的多尺度特征 (如尖锐的窄峰和宽缓的峰包)的捕捉能力有待进一步验证。此外,还有一些研究尝试对复杂的混合光谱信号进行解耦,来分离光谱中相关的特征和其他无关变量,为理解材料老化过程中的微观化学变化机理提供了强大的定量工具[22],但这些解混模型通常需要更复杂的先验知识,且主要目标是成分分析而非系统状态的直接评估。考虑到对于变压器绝缘状态快速诊断这一特定应用场景,本文旨在探索一种能够有效融合多尺度特征信息的网络结构,以期在计算成本和模型性能之间取得更优的平衡。
综上,本文通过加热箱对油纸绝缘进行加速老化,获取不同程度的老化样本,使用拉曼光谱仪对各个时间段的老化样本进行检测,采用S-G平滑法和airPLS法对拉曼光谱信号进行降噪和去基线,使用基于Inception-ResNet模块融合空洞卷积的1D-CNN算法搭建判别模型,对油纸绝缘老化程度进行分类,最后,将此模型评估效果与其他传统机器学习进行对比,研究本文算法的优越性。
一维卷积神经网络 (1D-CNN)是为一维数据而设计的深度学习架构,其核心结构是由卷积层、池化层和全连接层等多层级联的模块组成。
其中,卷积层主要用于提取序列的特征,与二维卷积神经网络不同的是,1D-CNN是采用单通道一维卷积核沿着输入数据进行移动,如图1所示,设卷积核的大小为k,权重为W=[ω1,ω2,···,ωk],偏置为b,则卷积操作的表达式为
$ \boldsymbol{Y}=f\left (\boldsymbol{X}*\boldsymbol{W}+\boldsymbol{b}\right)\;, $
式中:X为输入数据向量;*表示卷积操作;f (·)为该卷积层的激活函数。对于输入序列X的每个位置i,卷积输出为
$ \boldsymbol{y}_i=\sum_{j=0}^{k-1}x_{i+j}\cdot\omega_j+\boldsymbol{b}\; . $
卷积核每次移动都会与覆盖上的数据进行卷积形成一个新的数字,当对整个输入数据依次进行卷积操作后,会形成一个新的特征图谱,进行初级特征捕获。为了帮助网络更好地处理非线性问题,卷积操作后通常会连接激活函数来增强网络的非线性能力,本文采用ReLU函数作为激活函数。
池化层的作用是在保留重要特征的同时对卷积层输出的特征图谱进行压缩,在减少网络参数量和计算量的同时降低模型过拟合的风险,提高模型的整体效率。目前,最常用的池化方法有最大池化和平均池化,其数学表达式分别为
$ q_{{i}}=\max \left (y_{{i} \cdot {s}}, y_{{i} \cdot {s}+1}, \ldots, y_{{i} \cdot {s}+{p}-1}\right), $
$ q_i=\frac{1}{p} \sum_{j=0}^{p-1} y_{{i} \cdot {s}+{j}} ,$
式中:p为池化窗口大小;y为池化窗口内的数据;s为步长。
全连接层通常处于网络末端,负责将前面卷积层和池化层提取的特征进行铺平整合并与样本标记空间映射,进行最终的分类或回归。一维卷积神经网络结构图如图2所示。
一维空洞卷积是一维卷积的一种变体,其原理是通过引入膨胀率 (dilation rate)这个超参数,在卷积核的元素间插入空洞,在不增加卷积计算参数量的情况下增大卷积核能够覆盖的区域,使得模型能够捕捉到更广泛的上下文信息,对于需要考虑长距离依赖关系的任务非常重要。一维空洞卷积的示意图如图3所示,卷积核大小为1×3,膨胀率d分别为1、2和3。
由于空洞卷积的特征提取方式单一,容易导致丢失特征,为了更好地提取拉曼光谱数据中的局部细节和全局上下文信息,要求模型具有多尺度接受域。Inception 网络是一种高效的卷积神经网络结构,其可以通过并行使用不同大小的卷积核来捕获数据中不同尺度的特征,使得网络能够同时捕捉局部细节和全局信息,而空洞卷积可以在不增加计算成本的情况下扩展感受域,从而提高模型的特征表示能力[23]。为避免随着网络深度的增加,在训练过程中产生梯度消失问题,在模型中引入残差连接,融合多尺度特征提取与残差学习机制,如图4所示,其结构包含四个并行特征处理分支和一个自适应残差通路。
其中,分支1通过点卷积实现通道维度的线性变换,用于捕捉局部特征的通道相关性;分支2通过“压缩-扩展”策略,先以1×1卷积减少计算量,再通过3×1空洞卷积捕获中等范围时序模式;分支3同样先以1×1卷积减少计算量,但使用5×1空洞卷积核提取长程依赖特征;分支4通过最大池化保留显著特征后,利用1×1卷积调整通道维度,保留原始特征的鲁棒统计量。
为缓解深层网络梯度消失问题,引入自适应残差连接:
$ {\boldsymbol{F}}_{\text{out}}=\text{Concat}\left ({\boldsymbol{B}}_{1},{\boldsymbol{B}}_{2},{\boldsymbol{B}}_{3},{\boldsymbol{B}}_{4}\right)+\rho \left ({\boldsymbol{F}}_{\text{in}}\right) \;,$
式中:B1B2B3B4分别为四个独立的特征处理分支的输出;$ \rho ({{\boldsymbol{F}}}_{{\mathrm{in}}}) $为维度适配函数。
$ \rho\left ({\boldsymbol{F}}_{\text {in }}\right)=\left\{\begin{array}{ll}{\boldsymbol{F}}_{\text {in }} & C_{\text {in }}=C_{\text {out }} \\\operatorname{Conv} 1 {\mathrm{D}}_{1 \times 1}\left ({\boldsymbol{F}}_{\text {in}},{ \boldsymbol{W}}_{\mathrm{r}}\right) & {\text { otherwise }}\end{array}\right.. $
当输入的通道数与输出通道数不相等时,通过可学习的1×1卷积核实现通道维度对齐,确保残差项与主路径输出的逐元素相加可行性。
将空洞Inception-ResNet模块接入1D-CNN模型中,整体模型的架构如图5所示,输入数据首先经过一个标准的一维卷积层 (Conv1d),其主要目的是对原始输入序列进行初步的特征映射。在经过初步特征提取后,应用一个一维最大池化层 (MaxPool1d)对卷积后的特征图进行下采样,剔除不重要的冗余特征,减少后续的计算量并提高模型的平移不变性。池化后的特征图被送入两层的空洞Inception-ResNet模块中,提取多尺度的特征表达,提取特征后再次通过一个最大池化层,以进一步降低特征图的维度,保留其中的重要的高级特征。最后将经过多层卷积与池化操作后生成的多维特征图进行扁平化处理,转换为一个一维向量后输入至全连接层中,进行最终的分类判决。
本研究采用矿物绝缘油和牛皮绝缘纸,将绝缘油与绝缘纸按10∶1的比例进行混合,将其放置于加热箱内,并将加热箱温度精确控制在120 ℃进行恒温加热,使其加速老化[24],每隔24 h取样一次,每次采集15个平行样本,共采取加热周期为480 h (20天)的300个老化样品。在此温度环境下,随着加热时间的逐步延长,变压器油的老化也逐渐加剧,不同老化程度的样本如图6所示,从左往右老化程度依次加剧,可以看到随着老化程度的加剧,变压器油的颜色也越来越深。
糠醛作为辨别油纸绝缘老化的常用方法,可以很好地表明变压器当前的老化程度,随着加热时间的延长,油中的糠醛含量也在增高。因此,采用气相色谱方法对所取样本的糠醛含量进行实测,按所测的糠醛含量把样品的老化程度分为四类,如表1所示。
本实验采用深圳市依迈光电有限公司生产的RA200手持式拉曼光谱仪器对经过不同加热时间处理的变压器油样进行拉曼光谱检测,选择该型号的主要依据在于其高便携性、良好的信噪比,符合本研究对绝缘油状态进行快速、无损评估的应用背景。检测平台的结构如图7所示。
设置激光器的功率大小为300 mW,激发光波长为784.711 nm,积分时间为500 ms,不同老化阶段的绝缘油测试结果如图8所示,观察样本的拉曼光谱图可以发现,在拉曼频移为1500 cm−1和3000 cm−1的这两个地方附近有着明显的谱峰,且随着老化时间的延长,基线也随之升高。因此,为避免噪声和荧光信号对后续判别结果影响,需要先对原始拉曼光谱进行预处理,再进行后续分析。
采用S-G平滑法和airPLS对拉曼光谱原始信号进行降噪和基线去除的预处理。S-G平滑法是一种基于局部多项式拟合的数字信号降噪技术。其核心原理是通过在一个滑动窗口内对数据点进行多项式拟合,从而估算出平滑后的信号值。
自适应迭代重加权惩罚最小二乘法 (Adaptive iteratively reweighted penalized least squares, airPLS)是一种用于光谱数据基线校正的算法,适用于拉曼光谱中荧光背景去除[25]。其原理基于惩罚最小二乘法 (penalized least squares, PLS),并通过迭代加权的方式优化基线拟合效果。
为有效提取拉曼光谱中的特征信息,防止含有大量噪声的信号对基线拟合效果的影响,本研究采用串行处理的方式对原始光谱进行预处理。第一步,首先采用S-G平滑法对原始信号进行降噪处理,以消除高频噪声的干扰。第二步,在降噪后的光谱基础上,再利用airPLS算法进行基线校正,以去除荧光背景等因素造成的基线漂移,获得高质量的光谱数据用于后续模型训练。
通过降噪和去基线的预处理,拉曼光谱如图9所示。
图9可以看出,经过降噪和去基线后,特征峰强度与峰形得到显著优化,基线趋于平缓,为后续模型训练提供高质量输入数据。
在模型训练之前需要对其结构参数进行设置,以保障模型的性能,设置初始特征提取层中的卷积核通道数量为64,选择两倍最大池化,采用交叉熵损失函数 (cross-entropy loss)作为模型训练过程中目标函数,用于评估模型的预测输出与真实标记之间的偏差。为降低在有限样本集上训练复杂网络可能带来的过拟合风险,我们采用了AdamW优化器对整个网络的权重进行迭代更新以使损失函数趋于最小,并设置了权重衰减 (weight decay)参数,这相当于对模型权重施加了L2正则化,有效抑制了模型复杂度的过度增长,模型结构中使用的最大池化层 (MaxPool)本身也具有一定的正则化效果,增强了模型对输入微小变化的鲁棒性。
对空洞Inception-ResNet模块中四个分支的参数设置如表2所示。
将300个原始拉曼光谱老化样本数据按老化程度进行标记后输入到设置好的模型中,模型的整体工作流程如图10所示。
在拉曼光谱数据经过平滑以及去荧光背景等预处理后,按8∶2的比例随机划分成训练集和测试集,将训练集输入到基于空洞Inception-ResNet模块的1D-CNN老化判别模型中进行迭代训练,训练结束后将测试集输入最终模型进行测试,根据模型输出验证模型性能。
为全面评估所提模型框架的有效性,首先将原始的1D-CNN模型与两种在光谱数据分类任务中被广泛应用的机器学习模型—K近邻 (KNN)分类器和支持向量机 (SVM)分类器进行性能比较。为确保公正性,实验采用重复随机抽样验证的策略:通过改变随机数种子,随机挑选不同的训练集和测试集组合,进行10次独立重复的判别实验,以减少单次数据划分带来的偶然性偏差,从而更可靠地评估各模型的泛化能力。得到的结果如图11所示。
图11直观展示了这三种算法在10次独立实验中准确率波动的情况。可以看出,尽管各算法的性能在不同数据划分下均表现出一定的波动,但1D-CNN老化判别模型在10轮测试结果中有7轮准确率大于或等于另外两种算法,整体表现更为出色。
为进一步评估模型性能,引入平均准确率、准确率标准差和召回率对模型的检测结果进行综合评估,平均准确率反映了在多次实验中的整体性能;准确率标准差则衡量了模型性能的稳定性,标准差越小,表明模型对数据划分的敏感性越低,性能越稳定;召回率则关注模型正确识别出所有相关类别的能力。三个诊断模型的判别结果对比如表3所示。
表3可知,原始1D-CNN 模型在平均准确率 (89.83%)和平均召回率 (89.26%)两项关键指标上均高于KNN (平均准确率87.50%,召回率85.85%)和 SVM (平均准确率87.50%,召回率86.43%)这两个模型,这表明即使是基础的深度学习模型,在从高维、复杂的拉曼光谱数据中自动学习和提取特征方面也具有超越传统机器学习方法的先天优势,这是因为传统方法往往需要手工设计特征或依赖于特定的数据变换,而深度学习能够直接从原始光谱中学习深层的特征表示。此外,1D-CNN的准确率标准差 (2.54%)低于KNN (4.92%)和SVM (3.79%),这意味着 1D-CNN 模型在不同数据组合上的表现更为一致,说明其具有更强的稳定性与可靠性。
此对比结果初步证实基于一维卷积神经网络的方法在处理光谱老化评估任务中的潜力,并为后续引入更复杂的空洞卷积Inception-ResNet模块提供理论依据。
本节核心在于验证所提出的融合空洞卷积Inception-ResNet 模块的1D-CNN (以下简称“本文模型”)相对于原始1D-CNN在油纸绝缘老化诊断任务上的性能提升。设置训练的总轮次 (epochs)为 50,学习率为0.0001,对本文模型进行迭代训练,训练过程中的损失值 (Loss)变化曲线如图12所示。
通过训练损失图可以看到,在训练的前10轮次,模型的损失函数值快速下降,这表明模型能够快速从训练集的样本中学习到有效特征,大约在第20轮次的位置,损失值逐渐收敛到一个较低的水平,并在之后的训练过程中保持着平稳的状态。由于自适应残差连接的存在,确保梯度能够顺畅地反向传播至网络的较浅层,避免因网络加深而导致的梯度信号衰减,从而使得整个深度网络都能得到有效的训练。如果存在严重的梯度消失问题,我们预期会观察到损失长时间停滞不前或收敛极为缓慢的现象,而图中平稳高效的收敛过程佐证了本文模型设计的有效性。训练完成后,使用测试集分别在本文模型和同样条件训练好的原始1D-CNN、Inception-1DCNN模型中进行老化状态判别,三种判别模型分类结果的混淆矩阵如图13所示。
混淆矩阵直观地展示了模型对各个老化阶段的预测准确情况。由图13(a)可知,在同一数据条件下,原始1D-CNN模型在60个测试样本中总共错误划分了6个样本,且在各个老化阶段均有样本被错误地划分。相比之下,图13(b)中Inception-1DCNN模型仅有4个样本被错误划分,而图13(c)显示本文模型在Inception-1DCNN模型的基础上准确率再次提升,老化判别模型在60个测试集中仅有2个样本被错误划分,且主要集中在区分度可能较低的老化前中期的样本上,其余样本老化类别均被正确预测,因此,在单次测试中,本文模型的整体测试准确率达到了96.67%,相较于原始1D-CNN (准确率90%),Inception-1DCNN模型 (准确率93.33%),本文模型的绝对准确率依次提升了6.67%和3.34%。为避免单次实验的偶然性和随机性,采用不同的随机数种子来生成不同的训练集和测试集,对三种算法进行5次独立实验,并加入召回率和F1分数进行评估,5次实验的结果如表4所示。同时,为对计算开销与推理效率进行定量分析,引入参数量、FLOPs、单样本推理时间以及总训练时长这四个参数对几种算法的计算成本进行比较,如表5所示。
表4可以发现,在5次实验中本文模型的三项指标均优于另外两种模型,这证明了本文模型所使用的空洞卷积Inception-ResNet模块所带来的优势并非偶然。相较于原始1D-CNN,使用Inception模块可以通过不同的卷积核提取数据中的不同特征信息,同时捕捉局部细节和全局信息,对比使用单一的卷积核有着明显的优势。相较于原始的Inception模块,本文通过引入空洞卷积在不显著增加计算成本的前提下有效提升卷积核的感受野,同时加入残差网络又有效缓解了深层网络的梯度消失或爆炸问题,解决了网络退化现象,正是这些结构上的优化协同作用,使得本文模型能够更有效地学习和表征拉曼光谱中的老化特征,从而获得了更高的分类准确率和更稳定的性能表现。而由表5可以发现,由于其采用了多分支结构以捕获多尺度特征以及残差的使用,本文提出的模型在计算成本与推理效率上相较于原始1D-CNN和Inception-1DCNN有所增加,然而,这种复杂度的增加是适度的,并且单样本推理时间仍处于较低水平 (0.1424 ms),完全满足实时监测的需求。考虑到本文模型在准确率上带来的显著提升,这种计算成本的增加是合理且富有成效的。
综合上述实验结果与对比分析,本文提出的基于空洞 Inception-ResNet 模块的改进型 1D-CNN 模型,在变压器油纸绝缘老化状态的智能诊断任务中展现出卓越的性能。然而,在实际变压器老化评估的分类任务中,数据库的老化数据样本并不总是平衡的,它会对模型性能产生重要影响,具体而言,当模型在不平衡数据集上训练时,分类器会倾向于过多地关注样本量大的“多数类”,而忽视样本量少的“少数类”,这种偏见是极其危险的,因为它可能导致模型无法准确识别出稀有但至关重要的严重老化样本 (通常对应少数类)。因此,为进一步验证模型的鲁棒性和泛化能力,验证其在不平衡样本及小样本中的性能,通过使用不同数量的样本子集,以及改变训练集和测试集的划分比例对模型进行进一步实验。将300个样本随机组合分为A、B、C三组具有不同样本量的子集,组合结果如表6所示。
将三组样本子集分别按照9∶1、8.5∶1.5、8∶2、7∶3这四种不同的比例划分训练集和测试集,分别对本文模型进行训练和评估,通过多次实验后实验结果如表7所示。
表7的结果可以看出,本文提出的模型在不同规模的数据集 (A组、B组、C组)以及不同的训练集与测试集划分比例下,均表现出优异且稳定的性能。具体而言,对于样本量最大且数据平衡的A组 (300个样本),模型在所有划分比例下均取得了超过93%的平均准确率,最高可达95.33% (8∶2及9∶1划分),F1分数也相应地保持在0.93以上。即使在样本分布不平衡的B组 (230个样本),本文提出的模型依然取得了普遍高于92.5%的平均准确率,且各项指标并未出现显著恶化,这表明本模型通过其多尺度特征提取能力,成功学习到了各个老化阶段的内在、本质的光谱特征,而非仅仅依赖于类别样本数量的统计优势。而对于样本量较少的C组 (170个样本),模型的平均准确率依然保持在较高水平 (91.77%),说明本文模型在样本量有限的情况下,依然能够学习到样本的老化特征。从而维持良好的分类效果,各项评价指标均保持在较高水平。因此,模型展现出了对数据不平衡问题和样本量较小问题的良好鲁棒性,这对于其在实际应用中的可靠性至关重要。
本文结合拉曼光谱技术构建了一种融合空洞卷积、多分支Inception结构和自适应残差连接的改进型1D-CNN油纸绝缘老化评估模型,成功解决了传统检测方法周期长、破坏性、精度不足以及常规机器学习模型处理高维光谱数据时效率低、泛化能力弱等问题。通过热老化实验制备样本并利用拉曼光谱仪进行检测,结合S-G平滑与airPLS算法对光谱数据进行降噪和基线校正预处理,有效提升光谱数据质量和特征峰清晰度,为后续精准评估奠定了坚实基础。同时,为进一步提高评估精度,解决深层网络的梯度消失及单一尺度特征核提取特征具有局限性的问题,引入融合空洞卷积的Inception-ResNet模块,凭借其多分支并行空洞卷积对多尺度特征的提取能力以及自适应残差连接对梯度消失的缓解作用,在油纸绝缘老化状态的识别上展现出卓越性能。实验结果证明,所提出的模型与传统的K近邻 (KNN)、支持向量机 (SVM)算法以及原始的1D-CNN和Inception-1DCNN模型相比,在测试集上实现了高达96.67%的分类准确率,并在准确率、召回率及F1分数等多项评价指标上均显著优于对比模型。进一步的鲁棒性与泛化能力测试表明,即便在不同样本数量、不平衡数据集及不同训练测试集划分比例的条件下,模型依然能保持高效且稳定的诊断性能,为变压器绝缘状态的可靠评估和安全运行提供更可靠的技术支撑。
尽管本文模型取得了良好的效果,但仍存在一定的局限性:首先,模型在模拟热老化的实验室条件下训练,其在真实变压器复杂工况 (如多因素老化、强噪声干扰)下的性能有待验证;其次,与所有深度学习模型类似,本文模型是一个‘黑箱’,其决策依据 (如关键谱峰)尚不明确,可解释性有待加强。因此,提升模型在复杂工况下的鲁棒性并增强其可解释性将是我们未来研究的核心方向,在未来的工作中可进一步探索将更加先进高效的模型结构融入本模型框架,以增强对关键老化特征的识别能力。此外,与基于Transformer、LSTM等更前沿的模型在更大规模数据集上进行对比,也将是进一步验证本模型性能的重要方向。

参考文献 引证文献
排序方式:
1
梁栋, 朱建华, 张翠, 等. 变压器状态评估及故障诊断研究综述[J]. 变压器, 2024, 61 (2): 35−43.
Liang D, Zhu J H, Zhang C, et al. Review of transformer condition assessment and fault diagnosis[J]. Transformer, 2024, 61 (2): 35−43.
2
高浩, 刘庆珍, 蔡金锭. 基于去极化电流Prony拟合的油纸绝缘德拜参数辨识方法[J]. 高压电器, 2020, 56 (11): 210−218.
Gao H, Liu Q Z, Cai J D. Debye parameter identification method of oil-paper insulation based on depolarization current Prony algorithm fitting[J]. High Voltage Appar, 2020, 56 (11): 210−218.
3
陈啸轩, 邹阳, 翁祖辰, 等. 基于IKNN和LOF的变压器回复电压数据清洗方法研究[J]. 电子测量与仪器学报, 2024, 38 (2): 92−100.
Chen X X, Zou Y, Weng Z C, et al. Recovery voltage data cleaning method for transformer based on IKNN and LOF[J]. J Electron Meas Instrum, 2024, 38 (2): 92−100.
4
陈钰林, 许再尧, 莫元雄, 等. 基于绝缘纸聚合度预测的油浸式变压器老化评估方法研究[J]. 红水河, 2023, 42 (5): 107−111.
Chen Y L, Xu Z Y, Mo Y X, et al. Aging evaluating method for oil immersed transformer based on insulating paper polymerization degree prediction[J]. Hongshui River, 2023, 42 (5): 107−111.
5
李杰, 周渠, 贾路芬, 等. 红外、拉曼光谱的变压器油中糠醛检测方法对比研究[J]. 光谱学与光谱分析, 2024, 44 (1): 125−133.
Li J, Zhou Q, Jia L F, et al. Comparative study on detection methods of furfural in transformer oil based on IR and Raman spectroscopy[J]. Spectrosc Spectral Anal, 2024, 44 (1): 125−133.
6
王加安, 刘立人, 李延, 等. 基于拉曼高光谱成像技术检测面粉中的偶氮甲酰胺[J]. 电子测量技术, 2022, 45 (14): 97−102.
Wang J A, Liu L R, Li Y, et al. Detection of azoformamide in flour based on Raman hyperspectral imaging[J]. Electron Meas Technol, 2022, 45 (14): 97−102.
7
叶轲夫, 谢敏捷, 陈兴祺, 等. 拉曼光谱技术在环境微纳塑料检测中的应用与挑战[J]. 化学进展, 2025, 37 (1): 2−15.
Ye K F, Xie M J, Chen X Q, et al. Raman spectroscopy in the detection of environmental micro-and nanoplastics: applications and challenges[J]. Prog Chem, 2025, 37 (1): 2−15.
8
Guo N Z, Niu S J, Geng Y, et al. Non-destructive quantification of low colchicine concentrations in commercially available tablets using transmission Raman spectroscopy with partial least squares[J]. Int J Pharm: X, 2025, 9: 100321.
9
杨定坤. 油纸绝缘老化拉曼光谱多尺度特征提取及集成增强神经网络诊断研究[D]. 重庆: 重庆大学, 2021. https://doi.org/10.27670/d.cnki.gcqdu.2021.000468.
Yang D K. Study on multi scale feature extraction and integrated enhanced neural network diagnosis of oil-paper insulation aging by Raman spectroscopy[D]. Chongqing: Chongqing University, 2021. https://doi.org/10.27670/d.cnki.gcqdu.2021.000468.
10
陈新岗, 张文轩, 范益杰, 等. 基于局部线性嵌入的油纸绝缘拉曼光谱老化状态判别[J]. 激光与光电子学进展, 2025, 62 (3): 0330003.
Chen X G, Zhang W X, Fan Y J, et al. Determination of aging state of oil-paper insulation Raman spectrum based on local linear embedding[J]. Laser Optoelectron Prog, 2025, 62 (3): 0330003.
11
Fang J, Lin X, Zhou F X, et al. Site assessment of transformer state based on individual Raman spectrum equipment[J]. J Phys Conf Ser, 2023, 2584 (1): 012072.
12
陈新岗, 范益杰, 马志鹏, 等. 基于集成增强KNN的油纸绝缘原始拉曼光谱老化状态判别[J]. 激光与光电子学进展, 2023, 60 (21): 2130002.
Chen X G, Fan Y J, Ma Z P, et al. Aging state discrimination of oil-paper insulation using raman spectroscopy and integrated enhanced KNN[J]. Laser Optoelectron Prog, 2023, 60 (21): 2130002.
13
汪帮瑞, 薛建侠, 左小玉, 等. 基于拉曼光谱评估变压器油-屏障式绝缘老化状态[J]. 电气工程, 2023, 11 (2): 64−73.
Wang B R, Xue J X, Zuo X Y, et al. Evaluation of transformer oil-barrier based on Raman spectroscopy insulation aging state[J]. J Electr Eng, 2023, 11 (2): 64−73.
14
郝明, 白鹤, 徐婷婷. 融合ResNeSt和多尺度特征融合的遥感影像道路提取[J]. 光电工程, 2025, 52 (1): 240236.
Hao M, Bai H, Xu T T. Remote sensing image road extraction by integrating ResNeSt and multi-scale feature fusion[J]. Opto-Electron Eng, 2025, 52 (1): 240236.
15
张晨晨, 王帅, 王文一, 等. 针对人脸识别卷积神经网络的局部背景区域对抗攻击[J]. 光电工程, 2023, 50 (1): 220266.
Zhang C C, Wang S, Wang W Y, et al. Adversarial background attacks in a limited area for CNN based face recognition[J]. Opto-Electron Eng, 2023, 50 (1): 220266.
16
Song R M, Chen W G, Yang D K, et al. Aging assessment of oil-paper insulation based on visional recognition of the dimensional expanded Raman spectra[J]. IEEE Trans Instrum Meas, 2021, 70: 6007110.
17
钟明杉, 李兆飞, 张奕杰, 等. 基于一维卷积神经网络的天然气管道泄漏检测模型[J]. 国外电子测量技术, 2023, 42 (5): 62−68.
Zhong M S, Li Z F, Zhang Y J, et al. Natural gas pipeline leakage detection model based on 1-dimensional convolutional neural network[J]. For Electron Meas Technol, 2023, 42 (5): 62−68.
18
伍济钢, 文港, 杨康. 改进一维卷积神经网络的航空发动机故障诊断方法[J]. 电子测量与仪器学报, 2023, 37 (3): 179−186.
Wu J G, Wen G, Yang K. Improved one-dimensional convolutional neural network for aero-engine fault diagnosis[J]. J Electron Meas Instrum, 2023, 37 (3): 179−186.
19
Xiong S Y, Wang C X, Zhu C S, et al. Dual detection of urea and glucose in sweat using a portable microfluidic SERS sensor with silver nano-tripods and 1D-CNN model analysis[J]. ACS Appl Mater Interfaces, 2024, 16 (48): 65918−65926.
20
谢宇浩, 董前民, 金尚忠, 等. 基于深度神经网络的危险化学品拉曼光谱识别[J]. 激光与光电子学进展, 2025, 62 (5): 0530002.
Xie Y H, Dong Q M, Jin S Z, et al. Raman spectroscopic identification of hazardous chemicals based on a deep neural network[J]. Laser Optoelectron Prog, 2025, 62 (5): 0530002.
21
Xiong C C, Zhong Q S, Yan D H, et al. Multi-branch attention Raman network and surface-enhanced Raman spectroscopy for the classification of neurological disorders[J]. Biomed Opt Express, 2024, 15 (6): 3523−3540.
22
Georgiev D, Fernández-Galiana Á, Pedersen S V, et al. Hyperspectral unmixing for Raman spectroscopy via physics-constrained autoencoders[J]. Proc Natl Acad Sci USA, 2024, 121 (45): e2407439121.
23
涂潮, 刘万军, 赵琳琳, 等. 有限训练样本下的多尺度空洞密集网络高光谱影像分类[J]. 仪器仪表学报, 2024, 45 (4): 206−216.
Tu C, Liu W J, Zhao L L, et al. Multiscale dilated dense network for hyperspectral image classification with limited training samples[J]. Chin J Sci Instrum, 2024, 45 (4): 206−216.
24
李欢欢, 王思裕, 余武锟, 等. 绝缘纸在矿物油与菜籽油基天然酯中的加速热老化特性分析[J]. 绝缘材料, 2024, 57 (4): 44−48.
Li H H, Wang S Y, Yu W K, et al. Comparative analysis on accelerated thermal ageing characteristics of insulating paper in mineral oil and rapeseed oil based natural ester[J]. Insul Mater, 2024, 57 (4): 44−48.
25
郑成霞. airPLS算法去除拉曼光谱背景噪声的有效性研究[J]. 电子元器件与信息技术, 2021, 5 (2): 195−196.
Zheng C X. Research on the effectiveness of airPLS algorithm in removing background noise from Raman spectra[J]. Electron Compon Inf Technol, 2021, 5 (2): 195−196
2026年第53卷第4期
PDF下载
137
60
引用本文
BibTeX
文章信息
doi: 10.12086/oee.2026.250285
  • 接收时间:2025-09-23
  • 首发时间:2026-07-02
  • 出版时间:2026-04-24
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-23
  • 修回日期:2025-12-08
  • 录用日期:2025-12-09
基金
作者信息
    福州大学电气工程与自动化学院,福建 福州 350108

通讯作者:

宋福根,
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/oee/CN/10.12086/oee.2026.250285
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏