Structure of CRU.

<div><p>Insulator defect detection is a critical task in distribution network inspections. To address issues such as low detection accuracy, high model complexity, and large parameter counts caused by the variety of insulator defect types, this study propose a lightweight multi-defect de...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Weiyu Han (3448742) (author)
مؤلفون آخرون: Zixuan Cai (20765907) (author), Xin Li (51274) (author), Anan Ding (20765910) (author), Yuelin Zou (20765913) (author), Tianjun Wang (1524403) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1852022541122535424
author Weiyu Han (3448742)
author2 Zixuan Cai (20765907)
Xin Li (51274)
Anan Ding (20765910)
Yuelin Zou (20765913)
Tianjun Wang (1524403)
author2_role author
author
author
author
author
author_facet Weiyu Han (3448742)
Zixuan Cai (20765907)
Xin Li (51274)
Anan Ding (20765910)
Yuelin Zou (20765913)
Tianjun Wang (1524403)
author_role author
dc.creator.none.fl_str_mv Weiyu Han (3448742)
Zixuan Cai (20765907)
Xin Li (51274)
Anan Ding (20765910)
Yuelin Zou (20765913)
Tianjun Wang (1524403)
dc.date.none.fl_str_mv 2025-02-21T18:32:28Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0314225.g007
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/Structure_of_CRU_/28460469
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Medicine
Genetics
Molecular Biology
Sociology
Cancer
Space Science
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
suppress irrelevant features
siou loss function
simam attention mechanism
introducing scconv module
improve c2f module
higher average accuracy
experimental results show
encompasses four types
target detection dataset
low detection accuracy
improve detection accuracy
distribution network inspections
accelerate model convergence
insulator dataset compared
high model complexity
defect detection network
insulator defect types
defect detection
network improves
computational complexity
shedding insulator
model parameters
broken insulator
absent insulator
study propose
study creates
reduces spatial
lightweight algorithm
effective solution
critical task
channel redundancy
address issues
dc.title.none.fl_str_mv Structure of CRU.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>Insulator defect detection is a critical task in distribution network inspections. To address issues such as low detection accuracy, high model complexity, and large parameter counts caused by the variety of insulator defect types, this study propose a lightweight multi-defect detection network, LMD-YOLO, based on YOLOv8. The network improves the backbone by introducing SCConv module to improve C2f module, which reduces spatial and channel redundancy, lowering both computational complexity and the number of parameters. The SimAM attention mechanism is integrated to suppress irrelevant features and enhance feature extraction capabilities without adding extra parameters. The SIoU loss function is used in place of CIoU to accelerate model convergence and improve detection accuracy. Additionally, this study creates a target detection dataset that encompasses four types of insulators: insulator, absent insulator, broken insulator, and shedding insulator. Experimental results show that LMD-YOLO achieves a 2% higher average accuracy on the insulator dataset compared to YOLOv8n, with a 24.6% reduction in model parameters, offering an effective solution for smart grid inspections.</p></div>
eu_rights_str_mv openAccess
id Manara_477c4b2e1d6c8c4e97dcacd03d45cdf0
identifier_str_mv 10.1371/journal.pone.0314225.g007
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/28460469
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Structure of CRU.Weiyu Han (3448742)Zixuan Cai (20765907)Xin Li (51274)Anan Ding (20765910)Yuelin Zou (20765913)Tianjun Wang (1524403)MedicineGeneticsMolecular BiologySociologyCancerSpace ScienceBiological Sciences not elsewhere classifiedMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedsuppress irrelevant featuressiou loss functionsimam attention mechanismintroducing scconv moduleimprove c2f modulehigher average accuracyexperimental results showencompasses four typestarget detection datasetlow detection accuracyimprove detection accuracydistribution network inspectionsaccelerate model convergenceinsulator dataset comparedhigh model complexitydefect detection networkinsulator defect typesdefect detectionnetwork improvescomputational complexityshedding insulatormodel parametersbroken insulatorabsent insulatorstudy proposestudy createsreduces spatiallightweight algorithmeffective solutioncritical taskchannel redundancyaddress issues<div><p>Insulator defect detection is a critical task in distribution network inspections. To address issues such as low detection accuracy, high model complexity, and large parameter counts caused by the variety of insulator defect types, this study propose a lightweight multi-defect detection network, LMD-YOLO, based on YOLOv8. The network improves the backbone by introducing SCConv module to improve C2f module, which reduces spatial and channel redundancy, lowering both computational complexity and the number of parameters. The SimAM attention mechanism is integrated to suppress irrelevant features and enhance feature extraction capabilities without adding extra parameters. The SIoU loss function is used in place of CIoU to accelerate model convergence and improve detection accuracy. Additionally, this study creates a target detection dataset that encompasses four types of insulators: insulator, absent insulator, broken insulator, and shedding insulator. Experimental results show that LMD-YOLO achieves a 2% higher average accuracy on the insulator dataset compared to YOLOv8n, with a 24.6% reduction in model parameters, offering an effective solution for smart grid inspections.</p></div>2025-02-21T18:32:28ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0314225.g007https://figshare.com/articles/figure/Structure_of_CRU_/28460469CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/284604692025-02-21T18:32:28Z
spellingShingle Structure of CRU.
Weiyu Han (3448742)
Medicine
Genetics
Molecular Biology
Sociology
Cancer
Space Science
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
suppress irrelevant features
siou loss function
simam attention mechanism
introducing scconv module
improve c2f module
higher average accuracy
experimental results show
encompasses four types
target detection dataset
low detection accuracy
improve detection accuracy
distribution network inspections
accelerate model convergence
insulator dataset compared
high model complexity
defect detection network
insulator defect types
defect detection
network improves
computational complexity
shedding insulator
model parameters
broken insulator
absent insulator
study propose
study creates
reduces spatial
lightweight algorithm
effective solution
critical task
channel redundancy
address issues
status_str publishedVersion
title Structure of CRU.
title_full Structure of CRU.
title_fullStr Structure of CRU.
title_full_unstemmed Structure of CRU.
title_short Structure of CRU.
title_sort Structure of CRU.
topic Medicine
Genetics
Molecular Biology
Sociology
Cancer
Space Science
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
suppress irrelevant features
siou loss function
simam attention mechanism
introducing scconv module
improve c2f module
higher average accuracy
experimental results show
encompasses four types
target detection dataset
low detection accuracy
improve detection accuracy
distribution network inspections
accelerate model convergence
insulator dataset compared
high model complexity
defect detection network
insulator defect types
defect detection
network improves
computational complexity
shedding insulator
model parameters
broken insulator
absent insulator
study propose
study creates
reduces spatial
lightweight algorithm
effective solution
critical task
channel redundancy
address issues