PA-YOLO-Based Multifault Defect Detection Algorithm for PV Panels

IF 2.1 4区 工程技术 Q3 CHEMISTRY, PHYSICAL International Journal of Photoenergy Pub Date : 2024-02-08 DOI:10.1155/2024/6113260
Wang Yin, Zhao Jingyong, Xie Gang, Zhao Zhicheng, Hu Xiao
{"title":"PA-YOLO-Based Multifault Defect Detection Algorithm for PV Panels","authors":"Wang Yin, Zhao Jingyong, Xie Gang, Zhao Zhicheng, Hu Xiao","doi":"10.1155/2024/6113260","DOIUrl":null,"url":null,"abstract":"In recent years, solar photovoltaic (PV) energy, as a clean energy source, has received widespread attention and experienced rapid growth worldwide. However, the rapid growth of PV power deployment also brings important challenges to the maintenance of PV panels, and in order to solve this problem, this paper proposes an innovative algorithm based on PA-YOLO. First, we propose to use PA-YOLO’s asymptotic feature pyramid network (AFPN) instead of YOLOv7’s backbone network to support direct interactions of nonadjacent layers and avoid large semantic gaps between nonadjacent layers. For the occlusion problem of dense targets in the dataset, we introduce a repulsive loss function, which successfully reduces the occurrence of false detection situations. Finally, we propose a customized convolutional block equipped with an EMA mechanism to enhance the perceptual and expressive capabilities of the model. Experimental results on the dataset show that our proposed model achieves excellent performance with an average accuracy (mAP) of 94.5%, which is 6.8% higher than YOLOv7. In addition, our algorithm also succeeds in drastically reducing the model size from 71.3 MB to 48.4 MB, which well demonstrates the effectiveness of the model.","PeriodicalId":14195,"journal":{"name":"International Journal of Photoenergy","volume":null,"pages":null},"PeriodicalIF":2.1000,"publicationDate":"2024-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Photoenergy","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1155/2024/6113260","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"CHEMISTRY, PHYSICAL","Score":null,"Total":0}
引用次数: 0

Abstract

In recent years, solar photovoltaic (PV) energy, as a clean energy source, has received widespread attention and experienced rapid growth worldwide. However, the rapid growth of PV power deployment also brings important challenges to the maintenance of PV panels, and in order to solve this problem, this paper proposes an innovative algorithm based on PA-YOLO. First, we propose to use PA-YOLO’s asymptotic feature pyramid network (AFPN) instead of YOLOv7’s backbone network to support direct interactions of nonadjacent layers and avoid large semantic gaps between nonadjacent layers. For the occlusion problem of dense targets in the dataset, we introduce a repulsive loss function, which successfully reduces the occurrence of false detection situations. Finally, we propose a customized convolutional block equipped with an EMA mechanism to enhance the perceptual and expressive capabilities of the model. Experimental results on the dataset show that our proposed model achieves excellent performance with an average accuracy (mAP) of 94.5%, which is 6.8% higher than YOLOv7. In addition, our algorithm also succeeds in drastically reducing the model size from 71.3 MB to 48.4 MB, which well demonstrates the effectiveness of the model.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
基于 PA-YOLO 的光伏电池板多故障缺陷检测算法
近年来,太阳能光伏(PV)能源作为一种清洁能源,在全球范围内受到广泛关注并经历了快速增长。然而,光伏发电部署的快速增长也给光伏板的维护带来了重要挑战,为了解决这一问题,本文提出了一种基于 PA-YOLO 的创新算法。首先,我们建议使用 PA-YOLO 的渐近特征金字塔网络(AFPN)代替 YOLOv7 的骨干网络,以支持非相邻层的直接交互,避免非相邻层之间出现较大的语义差距。针对数据集中密集目标的遮挡问题,我们引入了排斥损失函数,成功减少了误检测情况的发生。最后,我们提出了一种配备 EMA 机制的定制卷积块,以增强模型的感知和表达能力。在数据集上的实验结果表明,我们提出的模型性能卓越,平均准确率(mAP)达到 94.5%,比 YOLOv7 高出 6.8%。 此外,我们的算法还成功地将模型大小从 71.3 MB 大幅减少到 48.4 MB,充分证明了模型的有效性。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
CiteScore
6.00
自引率
3.10%
发文量
128
审稿时长
3.6 months
期刊介绍: International Journal of Photoenergy is a peer-reviewed, open access journal that publishes original research articles as well as review articles in all areas of photoenergy. The journal consolidates research activities in photochemistry and solar energy utilization into a single and unique forum for discussing and sharing knowledge. The journal covers the following topics and applications: - Photocatalysis - Photostability and Toxicity of Drugs and UV-Photoprotection - Solar Energy - Artificial Light Harvesting Systems - Photomedicine - Photo Nanosystems - Nano Tools for Solar Energy and Photochemistry - Solar Chemistry - Photochromism - Organic Light-Emitting Diodes - PV Systems - Nano Structured Solar Cells
期刊最新文献
IGWO-VINC Algorithm Applied to MPPT Strategy for PV System Numerical Evaluation and Artificial Neural Network (ANN) Model of the Photovoltaic Thermal (PVT) System with Different Nanofluids Enhancing CsSn0.5Ge0.5I3 Perovskite Solar Cell Performance via Cu2O Hole Transport Layer Integration Investigation of the Performance of a Sb2S3-Based Solar Cell with a Hybrid Electron Transport Layer (h-ETL): A Simulation Approach Using SCAPS-1D Software Maximizing Conversion Efficiency: A Numerical Analysis on P+ a-SiC/i Interface/n-Si Heterojunction Solar Cells with AMPS-1D
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1