张引贤,张展耀,张希雅
ZHANG Yinxian,ZHANG Zhanyao,ZHANG Xiya

• 1:国网浙江省电力有限公司舟山供电公司
• 2:涿溪脑与智能研究所

摘要(Abstract)：

光伏表面异常检测对于光伏运维至关重要，而海岛光伏表面异常存在尺寸小、颜色差异小等难题。针对传统检测方法检测精度差和效率低的问题，提出一种基于边缘神经网络的海岛光伏表面异常检测方法。首先，结合卷积神经网络和注意力机制，构建多尺度特征融合的异常检测模型，深入挖掘细粒度异常特征，提升表面异常检测精度；进一步，采用双动态模型压缩技术，压缩冗余通道和特征块，显著降低模型计算复杂度，实现快速且高精度的异常检测。所提方法在舟山海岛光伏表面异常检测业务中表现良好，充分展示了其有效性和优越性。
Surface anomaly detection on photovoltaic(PV) panels is crucial for their operation and maintenance, especially in island environments where challenges such as small anomaly sizes and minimal color differences are prevalent. Due to the poor accuracy and low efficiency of existing detection methods, the paper proposes a surface anomaly detection method for island-based PV panels using edge neural networks. First, by use of convolutional neural networks(CNNs) and attention mechanisms, an anomaly detection model, characterized by multi-scale feature fusion, is constructed to explore the features of fine-grained anomalies, thereby enhancing the surface anomaly detection accuracy. Additionally, a dual dynamic model compression technique is employed to reduce redundant channels and feature blocks, significantly lowering the model's computational complexity and enabling rapid and accurate anomaly detection. The proposed method demonstrates strong performance in surface anomaly detection on PV panels in Zhoushan, highlighting its effectiveness and superiority.

关键词(KeyWords)： 海岛光伏;异常检测;模型压缩;多尺度特征融合
island-based PV panels;anomaly detection;model compression;multi-scale feature fusion

Abstract:

Keywords:

基金项目(Foundation): 浙江省自然科学基金资助项目（LDT23F01013F01）;; 国网浙江省电力有限公司科技项目（B311ZS230002）

作者(Author): 张引贤,张展耀,张希雅
ZHANG Yinxian,ZHANG Zhanyao,ZHANG Xiya

DOI: 10.19585/j.zjdl.202412010

参考文献(References)：

• [1]国网浙江省电力有限公司.浙江舟山公司：光伏项目并网助海岛“双碳”目标实现[J].农村电气化，2021(10):59.State Grid Zhejiang Electric Power Co., Ltd.. Zhejiang Zhoushan Company:Grid-connected photovoltaic projects help the island achieve the goal of “double carbon”[J].Rural Electrification,2021(10):59.
• [2]唐少南.极端天气影响的海岛微网群韧性评估与提升[D].吉林：东北电力大学，2022.TANG Shaonan.Resilience assessment and enhancement of island microgrids under extreme weather conditions[D].Jilin:Northeast Electric Power University,2023.
• [3]李民，杨暑森，李科锋，等.覆雪状态下光伏发电功率预测方法研究[J].高压电器，2023,59(9):250-257.LI Min,YANG Shusen,LI Kefeng,et al.Study on power prediction method of photovoltaic power generation under snow cover[J]. High Voltage Apparatus,2023,59(9):250-257.
• [4]王道累，姚从荣，李超，等.光伏组件热斑效应智能化检测的综述及展望[J].太阳能学报，2024,45(5):527-536.WANG Daolei,YAO Congrong,LI Chao,et al. Review and prospects of intelligent detection of hot spot effect of photovoltaic modules[J]. Acta Energiae Solaris Sinica,2024,45(5):527-536.
• [5]周全民，杜玉杰，叶征灯，等.YOLO系列算法在光伏组件缺陷检测中的应用与分析[J].新型工业化，2021,11(1):107-108.ZHOU Quanmin,DU Yujie,YE Zhengdeng,et al.Application and analysis of YOLO series algorithms in photovoltaic module defect detection[J].The Journal of New Industrialization,2021,11(1):107-108.
• [6]刘枭雄，郑茜颖.基于深度学习的光伏板缺陷分类定位算法研究[J].光电子技术，2024,44(1):54-60.LIU Xiaoxiong,ZHENG Xinying. Research on deep learning-based classification and localization algorithm for photovoltaic panel defects[J].Optoelectronic Technology,2024,44(1):54-60.
• [7] HE K M,ZHANG X Y,REN S Q,et al.Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR),27-30 Jun.2016,Las Vegas,NV,USA.IEEE,2016:770-778.
• [8] DOSOVITSKIY A,BEYER L,KOLESNIKOV A,et al.An image is worth 16x16 words:transformers for image recognition at scale[EB/OL].(2021-08-03)[2024-04-22].https：//arxiv.org/abs/2010.11929.
• [9] WANG A,CHEN H,LIN Z J,et al.RepViT:revisiting mobile CNN from ViT perspective[EB/OL].(2024-03-14)[2024-04-25].https：//arxiv.org/abs/2307.09283v8.
• [10] REN S,HE K,GIRSHICK R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149.
• [11]伊欣同，单亚峰.基于改进Faster R-CNN的光伏电池内部缺陷检测[J].电子测量与仪器学报，2021,35(1):40-47.YI Xintong,SHAN Yafeng.Photovoltaic cell internal defect detection based on improved Faster R-CNN[J].Journal of Electronic Measurement and Instrumentation,2021,35(1):40-47.
• [12]李琼，吴文宝，刘斌，等.基于迁移学习的光伏组件鸟粪覆盖检测[J].太阳能学报，2022,43(2):233-237.LI Qiong,WU Wenbao,LIU Bin,et al. Bird droppings coverage detection of photovoltaic module based on transfer learning[J].Acta Energiae Solaris Sinica,2022,43(2):233-237.
• [13] DENG J,DONG W,SOCHER R,et al. ImageNet:a large-scale hierarchical image database[C]//2009 IEEE Conference on Computer Vision and Pattern Recognition,20-25 Jun. 2009,Miami,FL,USA.IEEE,2009:248-255.
• [14]吕潇涵.基于改进Vision Transformer的光伏电池缺陷识别研究[J].计算技术与自动化，2023,42(4):33-40.LüXiaohan.Defect recognition of photovoltaic cells based on convolutional neural network and scaled vision transformer[J].Computing Technology and Automation,2023,42(4):33-40.
• [15] XIONG R B,YANG Y C,HE D,et al.On layer normalization in the transformer architecture[EB/OL].(2020-06-29)[2024-05-03].https：//arxiv.org/abs/2002.04745v2.
• [16]兰金江，曾学仁，方亮，等.基于无人机巡检的光伏缺陷检测与定位[J].科技创新与应用，2024,14(18):14-19.LAN Jinjiang,ZENG Xueren,FANG Liang,et al.Photovoltaic defect detection and location based on UAV patrol inspection[J]. Technology Innovation and Application,2024,14(18):14-19.
• [17]李峰，付敬，杨梅，等.基于监督学习的光伏组件检测[J].计量与测试技术，2024,51(6):1-3.LI Feng,FU Jing,YANG Mei,et al.Photovoltaic module detection based on supervised learning[J].Metrology and Measurement Technology,2024,51(6):1-3.
• [18] RUSSAKOVSKY O,DENG J,SU H,et al. ImageNet large scale visual recognition challenge[J]. International Journal of Computer Vision,2015,115(3):211-252.
• [19] ZHOU X Y,WANG D Q,PHILIPP KR?HENBüHL.Objects as points[EB/OL].(2019-04-25)[2024-05-03].https：//arxiv.org/abs/1904.07850.
• [20] REDMON J,DIVVALA S,GIRSHICK R,et al. You only look once:unified,real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR),27-30 Jun.2016,Las Vegas,NV,USA.IEEE,2016:779-788.
• [21]周得永，高龙琴.基于YOLOv3的太阳能电池板缺陷检测[J].南方农机，2022,53(4):142-144.ZHOU Deyong,GAO Longqin.Solar panel defect detection based on YOLOv3[J]. China Southern Agricultural Machinery,2022,53(4):142-144.
• [22] YU S X,CHEN T L,SHEN J Y,et al. Unified visual transformer compression[EB/OL].[2024-05-06].https：//arxiv.org/pdf/2203.08243.
• [23] WANG Z Y,LUO H,WANG P H,et al.Vtc-lfc:Vision transformer compression with low-frequency components[EB/OL].(2022-12-25)[2024-05-06]. https：//openreview.net/forum?id=HuiLIB6EaOk.
• [24]刘怀广，丁晚成，黄千稳.基于轻量化卷积神经网络的光伏电池片缺陷检测方法研究[J].应用光学，2022,43(1):87-94.LIU Huaiguang,DING Wancheng,HUANG Qianwen.Defects detection method of photovoltaic cells based on lightweight convolutional neural network[J]. Journal of Applied Optics,2022,43(1):87-94.