View in EDS HTML Full Text PDF Full Text

YOLO-MSG: A Lightweight and Real-Time Photovoltaic Defect Detection Algorithm for Edge Computing.

Saved in:
Bibliographic Details
Title: YOLO-MSG: A Lightweight and Real-Time Photovoltaic Defect Detection Algorithm for Edge Computing.
Authors: Zhu, Jingdong1 (AUTHOR), Qian, Xu1,2 (AUTHOR), Wang, Liangliang1,3 (AUTHOR), Yin, Chong1 (AUTHOR), Wang, Tao1 (AUTHOR), Xu, Zhanpeng1 (AUTHOR) xu_zp@hdec.com, Yao, Zhenqin2 (AUTHOR), Wang, Ban2,3 (AUTHOR)
Source: Energies (19961073). May2026, Vol. 19 Issue 9, p2043. 19p.
Subject Terms: *Edge computing, *Real-time computing, *Knowledge transfer, *Object recognition (Computer vision), *Artificial neural networks
Abstract: Photovoltaic (PV) power stations are pivotal for the renewable energy transition, yet their operational efficiency is often compromised by defects such as surface dust accumulation and cracks. Traditional manual inspections are labor-intensive and subjective, while conventional monitoring methods struggle with environmental interference. This study proposes YOLO-MSG, a lightweight framework specifically designed for the automated detection of PV module defects during system operation, including normal panels as well as defective conditions such as dusty and cracked panels. The methodology integrates a Multi-Scale Grouped Convolution (MSGC) module for enhanced feature extraction and a Group-Stem Decoupled Head (GSD-Head) to reduce parameter redundancy. Furthermore, a joint optimization strategy involving LAMP and logits-based knowledge distillation is employed to facilitate edge deployment. Experimental results on a specialized PV defect dataset demonstrate that YOLO-MSG achieves a superior balance between detection accuracy and computational cost. Compared to state-of-the-art models like YOLO11 and YOLOv12, YOLO-MSG significantly reduces GFLOPs and parameter count while maintaining highly competitive mean Average Precision (mAP), with improvements of 1.35% in mAP and 2.37% in mAP50-95 over the baseline models. Specifically, the model achieves an average inference speed of 90.30 FPS on the NVIDIA Jetson AGX platform. These findings confirm the algorithm's industrial viability, providing a robust and efficient solution for the real-time automated maintenance of photovoltaic infrastructures. [ABSTRACT FROM AUTHOR]
Database: Energy & Power Source
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: enr
DbLabel: Energy & Power Source
An: 193715939
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: YOLO-MSG: A Lightweight and Real-Time Photovoltaic Defect Detection Algorithm for Edge Computing.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Zhu%2C+Jingdong%22">Zhu, Jingdong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Qian%2C+Xu%22">Qian, Xu</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Liangliang%22">Wang, Liangliang</searchLink><relatesTo>1,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yin%2C+Chong%22">Yin, Chong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Tao%22">Wang, Tao</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Xu%2C+Zhanpeng%22">Xu, Zhanpeng</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> xu_zp@hdec.com</i><br /><searchLink fieldCode="AR" term="%22Yao%2C+Zhenqin%22">Yao, Zhenqin</searchLink><relatesTo>2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Ban%22">Wang, Ban</searchLink><relatesTo>2,3</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Energies+%2819961073%29%22">Energies (19961073)</searchLink>. May2026, Vol. 19 Issue 9, p2043. 19p.
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: *<searchLink fieldCode="DE" term="%22Edge+computing%22">Edge computing</searchLink><br />*<searchLink fieldCode="DE" term="%22Real-time+computing%22">Real-time computing</searchLink><br />*<searchLink fieldCode="DE" term="%22Knowledge+transfer%22">Knowledge transfer</searchLink><br />*<searchLink fieldCode="DE" term="%22Object+recognition+%28Computer+vision%29%22">Object recognition (Computer vision)</searchLink><br />*<searchLink fieldCode="DE" term="%22Artificial+neural+networks%22">Artificial neural networks</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Photovoltaic (PV) power stations are pivotal for the renewable energy transition, yet their operational efficiency is often compromised by defects such as surface dust accumulation and cracks. Traditional manual inspections are labor-intensive and subjective, while conventional monitoring methods struggle with environmental interference. This study proposes YOLO-MSG, a lightweight framework specifically designed for the automated detection of PV module defects during system operation, including normal panels as well as defective conditions such as dusty and cracked panels. The methodology integrates a Multi-Scale Grouped Convolution (MSGC) module for enhanced feature extraction and a Group-Stem Decoupled Head (GSD-Head) to reduce parameter redundancy. Furthermore, a joint optimization strategy involving LAMP and logits-based knowledge distillation is employed to facilitate edge deployment. Experimental results on a specialized PV defect dataset demonstrate that YOLO-MSG achieves a superior balance between detection accuracy and computational cost. Compared to state-of-the-art models like YOLO11 and YOLOv12, YOLO-MSG significantly reduces GFLOPs and parameter count while maintaining highly competitive mean Average Precision (mAP), with improvements of 1.35% in mAP and 2.37% in mAP50-95 over the baseline models. Specifically, the model achieves an average inference speed of 90.30 FPS on the NVIDIA Jetson AGX platform. These findings confirm the algorithm's industrial viability, providing a robust and efficient solution for the real-time automated maintenance of photovoltaic infrastructures. [ABSTRACT FROM AUTHOR]
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=enr&AN=193715939
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/en19092043
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 19
        StartPage: 2043
    Subjects:
      – SubjectFull: Edge computing
        Type: general
      – SubjectFull: Real-time computing
        Type: general
      – SubjectFull: Knowledge transfer
        Type: general
      – SubjectFull: Object recognition (Computer vision)
        Type: general
      – SubjectFull: Artificial neural networks
        Type: general
    Titles:
      – TitleFull: YOLO-MSG: A Lightweight and Real-Time Photovoltaic Defect Detection Algorithm for Edge Computing.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Zhu, Jingdong
      – PersonEntity:
          Name:
            NameFull: Qian, Xu
      – PersonEntity:
          Name:
            NameFull: Wang, Liangliang
      – PersonEntity:
          Name:
            NameFull: Yin, Chong
      – PersonEntity:
          Name:
            NameFull: Wang, Tao
      – PersonEntity:
          Name:
            NameFull: Xu, Zhanpeng
      – PersonEntity:
          Name:
            NameFull: Yao, Zhenqin
      – PersonEntity:
          Name:
            NameFull: Wang, Ban
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Text: May2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 19961073
          Numbering:
            – Type: volume
              Value: 19
            – Type: issue
              Value: 9
          Titles:
            – TitleFull: Energies (19961073)
              Type: main
ResultId 1