一种基于电致发光图像的无监督光伏电池异常检测方案

田家源; 王玉峰

doi:null

您当前的位置：

首页 >

文章列表页 >

一种基于电致发光图像的无监督光伏电池异常检测方案

更新时间：2026-03-03

- 一种基于电致发光图像的无监督光伏电池异常检测方案
- An Unsupervised Photovoltaic Module Anomaly Detection Scheme Based on EL Images
- 物联网学报 2026年
- 作者机构：
  
  1.南京邮电大学波特兰学院，江苏南京邮编210023
  2.南京邮电大学通信与信息工程学院，江苏南京邮编210023
- 作者简介：
  
  [ "田家源（2001- ），男，南京邮电大学硕士研究生，主要研究方向为深度学习，无监督异常检测。" ]
  [ "王玉峰（1974- ），男，博士，南京邮电大学通信与信息工程学院教授，博士生导师，主要研究数据科学以及人工智能（数据挖掘、深度学习和强化学习等）与算法机制设计研究和应用。" ]
- 基金信息：
- DOI：
  中图分类号： TP391.41
- 收稿：2025-07-29，
  
  修回：2025-10-20，
  
  录用：2025-10-20，
- 稿件说明：
移动端阅览
田家源, 王玉峰. 一种基于电致发光图像的无监督光伏电池异常检测方案[J/OL]. 物联网学报, 2026.

TIAN Jiayuan, WANG Yufeng. An Unsupervised Photovoltaic Module Anomaly Detection Scheme Based on EL Images[J/OL]. Chinese Journal on Internet of Things, 2026.
田家源, 王玉峰. 一种基于电致发光图像的无监督光伏电池异常检测方案[J/OL]. 物联网学报, 2026. DOI：

TIAN Jiayuan, WANG Yufeng. An Unsupervised Photovoltaic Module Anomaly Detection Scheme Based on EL Images[J/OL]. Chinese Journal on Internet of Things, 2026. DOI：

摘要

电致发光（EL）成像技术结合基于特征提取的监督学习方法常常被用于光伏电池的异常检测，但监督学习不仅标记成本高昂，且难以检测出训练集中没有的异常模式。无监督学习方法有助于解决上述问题，然而如何高效利用多尺度特征进行无监督异常检测仍缺乏成熟的范式。对此，本文提出了一种新颖的基于EL图像的无监督光伏电池异常检测方案—MFRAD。首先，利用深度卷积网络提取光伏电池EL图像的多尺度特征。其次，针对高维多尺度特征设计了高效的对抗重构模块。最后，综合特征空间和潜在空间中的重构误差组成异常评分，实现光伏电池异常的有效检测。实验结果表明，MFRAD在单晶和多晶光伏电池EL数据集上各取得了0.956，0.868的ROC-AUC，优于其他无监督异常检测方法。

Abstract

Electroluminescence (EL) imaging combined with feature-based supervised learning was often used for photovoltaic (PV) cell anomaly detection. However

supervised methods incurred high labeling costs and failed to detect anomaly patterns not present in the training set. Unsupervised learning was employed to address these issues. Nevertheless

no established paradigm had existed for the efficient utilization of multi-scale features in unsupervised anomaly detection. To tackle this challenge

a novel unsupervised PV cell anomaly detection method

MFRAD

was proposed. Multi-scale features of EL images were first extracted using a deep convolutional network. Then

an efficient adversarial reconstruction module was designed for high-dimensional multi-scale features. Finally

reconstruction errors in both the feature space and latent space were combined to generate anomaly scores

enabling effective detection of PV cell anomalies. The experimental results showed that MFRAD achieved ROC-AUCs of 0.956 and 0.868 on monocrystalline and polycrystalline photovoltaic cell EL datasets

outperforming other unsupervised anomaly detection methods.

关键词

Keywords

references

International Energy Agency . Renewables 2024 [R ] . Paris : IEA , 2024 . [2024-07-29 ] . Available: https://www.iea.org.

AKRAM M W , LI G , JIN Y , CHEN X , EL-BANBY G M . Failures of photovoltaic modules and their detection: A review [J ] . Applied Energy , 2022 , 313 : 118822 .

AL-OTUM H M . Deep learning-based automated defect classification in electroluminescence images of solar panels [J ] . Advanced Engineering Informatics , 2023 , 58 : 102147 .

OZTURK E , OGLIARI E , SAKWA M , DOLARA A , BLASUTTIGH N , PAVAN A M . Photovoltaic modules fault detection, power output, and parameter estimation: A deep learning approach based on electroluminescence images [J ] . Energy Conversion and Management , 2024 , 319 : 118866 .

Demirci M Y , Beşli N , Gümüşçü A . Efficient deep feature extraction and classification for identifying defective photovoltaic module cells in electroluminescence images [J ] . Expert Systems with Applications , 2021 , 175 : 114810 .

ZHANG J , YANG W , CHEN Y , DING M , HUANG H , WANG B , et al . Fast object detection of anomaly photovoltaic (PV) cells using deep neural networks [J ] . Applied Energy , 2024 , 372 : 123759 .

RAMADAN E A , MOAWAD N M , ABOUZALM B A , SAKR A A , ABOUZAID W F , EL-BANBY G M . An innovative transformer neural network for fault detection and classification for photovoltaic modules [J ] . Energy Conversion and Management , 2024 , 118 : 118718 .

NGUYEN H , NGUYEN T A , TOAN N D . Optimizing feature extraction and fusion for high-resolution defect detection in solar cells [J ] . Intelligent Systems with Applications , 2024 , 24 : 200443 .

DEMIRCI M Y , BEŞLI N , GÜMÜŞÇÜ A . An improved hybrid solar cell defect detection approach using Generative Adversarial Networks and weighted classification [J ] . Expert Systems with Applications , 2024 , 252 : 124230 .

XIA X , PAN X , LI N , et al . GAN-based anomaly detection: A review [J ] . Neurocomputing , 2022 , 493 : 497 – 535 .

OTAMENDI U , MARTINEZ I , OLAIZOLA I G , QUARTULLI M . A scalable framework for annotating photovoltaic cell defects in electroluminescence images [J ] . IEEE Transactions on Industrial Informatics , 2023 , 19 ( 9 ): 9361 – 9369 .

QIAN X , LI J , CAO J , WU Y , WANG W . Micro-cracks detection of solar cells surface via combining short-term and long-term deep features [J ] . Neural Networks , 2020 , 127 : 132 – 140 .

SCHLEGL T , SEEBÖCK P , WALDSTEIN S M , et al . f-AnoGAN: Fast unsupervised anomaly detection with generative adversarial networks [J ] . Medical Image Analysis , 2019 , 54 : 30 – 44 .

GULRAJANI I , AHMED F , ARJOVSKY M , DUMOULIN V , COURVILLE A . Improved training of Wasserstein GANs [C ] // Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS) . 2017 : 5767 – 5777 .

ZHU J , DENG F , ZHAO J , CHEN J . Adaptive aggregation-distillation autoencoder for unsupervised anomaly detection [J ] . Pattern Recognition , 2022 , 131 : 108897 .

HU D , ZHANG C , YANG T , FANG Q . A deep autoencoder with structured latent space for process monitoring and anomaly detection in coal-fired power units [J ] . Reliability Engineering & System Safety , 2025 , 261 : 111060 .

HADSELL R , CHOPRA S , LECUN Y . Dimensionality reduction by learning an invariant mapping [C ] // 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) . 2006 : 1735 – 1742 .

WANG Z , BOVIK A C , SHEIKH H R , SIMONCELLI E P . Image quality assessment: From error visibility to structural similarity [J ] . IEEE Transactions on Image Processing , 2004 , 13 ( 4 ): 600 – 612 .

RONNEBERGER O , FISCHER P , BROX T . U-Net: Convolutional networks for biomedical image segmentation [C ] // Proceedings of the Medical Image Computing and Computer-Assisted Intervention (MICCAI) . 2015 : 234 – 241 .

KANG X , LI Y , ZHANG Y , MA N , WEN L . Anomaly detection in concrete dam using memory-augmented autoencoder and generative adversarial network [J ] . Automation in Construction , 2024 , 168 ( A ): 105794 .

AKCAY S , ABARGHOUEI A A , BRECKON T P . GANomaly: Semi-supervised anomaly detection via adversarial training [C ] // Proceedings of the Asian Conference on Computer Vision (ACCV) . 2018 : 622 – 637 .

DENG H , LI X . Anomaly detection via reverse distillation from one-class embedding [C ] // Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . 2022 : 9727 – 9736 .

TIEN T D , DO H V , HUYNH T V , KIM D S . Revisiting reverse distillation for anomaly detection [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . 2023 : 24511 – 24520 .

WANG R , LI Z , ZHENG L , WANG W , LI S . Deep feature clustering for multi-class industrial image anomaly detection [J ] . Knowledge-Based Systems , 2025 , 311 : 113134 .

GUDOVSKIY D , ISHIZAKA S , KOZUKA K . CFLOW-AD: Real-time unsupervised anomaly detection with localization via conditional normalizing flows [C ] // Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) . 2022 : 1819 – 1828 .

DEFARD T , SETKOV A , LOESCH A , AUDIGIER R . PaDiM: A patch distribution modeling framework for anomaly detection and localization [C ] // ICPR International Workshops and Challenges . 2021 : 35 – 42 .

SHI Y , YANG J , QI Z . Unsupervised anomaly segmentation via deep feature reconstruction [J ] . Neurocomputing , 2021 , 424 : 9 – 22 .

TAO D , ZHANG W , MA Z , HOU Z , LU Z , ADAK C . Unsupervised anomaly detection for surface defects with dual-siamese network [J ] . IEEE Transactions on Industrial Informatics , 2022 , 18 ( 11 ): 7707 – 7717 .

SALEHI M , SADJADI N , BASELIZADEH S , ROHBAN M H , RABIEE H R . Multiresolution knowledge distillation for anomaly detection [C ] // Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . 2021 : 14897 – 14907 .

SIMONYAN K , ZISSERMAN A . Very deep convolutional networks for large-scale image recognition [EB/OL ] . ( 2014-09-04 ) [ 2024-07-29 ] . https://arxiv.org/abs/1409.1556 https://arxiv.org/abs/1409.1556 .

LI X , ZHENG Y , CHEN B , ZHENG E . Dual attention-based industrial surface defect detection with consistency loss [J ] . Sensors , 2022 , 22 ( 14 ): 5141 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据