基于D-DADA算法与DBE-YOLO网络的电表异常检测方法

doi:10.19678/j.issn.1000-3428.0070261

计算机工程 ›› 2026, Vol. 52 ›› Issue (5): 445-455. doi: 10.19678/j.issn.1000-3428.0070261

基于D-DADA算法与DBE-YOLO网络的电表异常检测方法

张蓬鹤¹, 杨艺宁¹^,*(), 王璧成¹, 易云齐², 唐忠瑞², 刘敏²

1. 中国电力科学研究院有限公司计量研究所, 北京 100192
2. 湖南大学电气与信息工程学院, 湖南长沙 410082

收稿日期:2024-08-16 修回日期:2024-10-16 出版日期:2026-05-15 发布日期:2025-01-07
通讯作者: 杨艺宁
作者简介:
张蓬鹤, 女, 教授级高级工程师、博士, 主研方向为电能计量、反窃电及低压用电安全技术
杨艺宁(通信作者), 工程师、硕士
王壁成, 工程师、硕士
易云齐, 硕士研究生
唐忠瑞, 硕士研究生
刘敏, 教授、博士
基金资助:
国家电网有限公司科技项目(5400-202355230A-1-1-ZN)

Anomaly Detection Method for Electricity Meter Based on D-DADA Algorithm and DBE-YOLO Network

ZHANG Penghe¹, YANG Yining¹^,*(), WANG Bicheng¹, YI Yunqi², TANG Zhongrui², LIU Min²

1. Measurement Research Institute, China Electric Power Research Institute, Beijing 100192, China
2. College of Electrical and Information Engineering, Hunan University, Changsha 410082, Hunan, China

Received:2024-08-16 Revised:2024-10-16 Online:2026-05-15 Published:2025-01-07
Contact: YANG Yining

摘要/Abstract

摘要：

目前用户侧智能电表的维护与检测主要依赖专业人员上门排查, 存在现场检验效率低、周期检定任务繁重、严重依赖人工经验等问题。基于电网巡检图片构建电表异常图像数据集, 针对电表图像背景复杂、目标尺寸不一、异常接线隐蔽遮挡等问题, 提出一种基于多样性感知的可微分自动数据增广(D-DADA)与双分支特征增强YOLO(DBE-YOLO)网络的电表异常检测方法。首先, 提出改进的DBE-YOLO网络, 通过引入级联空洞卷积增强模型对全局上下文信息与多尺度特征的提取, 设计双分支聚合结构弥补了原始模型感受野受限、卷积特征捕捉模式固定的缺陷。其次, 提出一种多样性感知的D-DADA算法, 设计了搜索策略多样性约束条件促进对更广泛数据增强策略的自动搜索, 从而帮助模型学习到不同场景、角度、光照等情况下的检测目标特征和模式, 解决数据类内差异较大导致的模型识别性能不足等问题。实验结果表明, 改进后的YOLOv8模型对8类电表异常的平均检测精度可达到79.6%, 相对于改进前提高了3.4百分点。

关键词: 电表, YOLOv8模型, 异常检测, DBE-C2f模块, 自动数据增广

Abstract:

Currently, the maintenance and anomaly detection of user-side smart meters primarily rely on professionals visiting the site, leading to low inspection efficiency, significant periodic testing burdens, and dependence on manual experience. A dataset of abnormal electricity meter images is created based on the inspection images obtained from a power grid company. This paper introduces a novel anomaly detection method for electricity meters that utilizes Diversity-Driven Differentiable Automatic Data Augmentation (D-DADA) algorithm and the Dual-Branch Feature Enhancement YOLO (DBE-YOLO) network to address issues such as complex backgrounds, varying target sizes, and obscured wiring in meter images. First, the DBE-YOLO model is designed to enhance the extraction of global contextual information and multiscale features by introducing cascaded dilated convolutions. It also employs a dual-branch aggregation network to overcome the limitations of the original model, including a restricted receptive field and fixed convolutional feature capture patterns. Second, the D-DADA algorithm is introduced, featuring a search strategy with diversity constraints to enhance the automatic discovery of a wider array of data augmentation strategies. This enables the model to learn the detection target features and patterns under various scenarios, angles, and lighting conditions, addressing the issue of insufficient model recognition performance owing to large intraclass variations. The experimental results indicate that the improved YOLOv8 model achieves an average detection accuracy of 79.6% across eight types of electricity meter anomalies, representing a 3.4 percentage point increase compared with the previous version.

Key words: electricity meter, YOLOv8 model, anomaly detection, DBE-C2f module, automatic data augmentation

张蓬鹤, 杨艺宁, 王璧成, 易云齐, 唐忠瑞, 刘敏. 基于D-DADA算法与DBE-YOLO网络的电表异常检测方法[J]. 计算机工程, 2026, 52(5): 445-455.

ZHANG Penghe, YANG Yining, WANG Bicheng, YI Yunqi, TANG Zhongrui, LIU Min. Anomaly Detection Method for Electricity Meter Based on D-DADA Algorithm and DBE-YOLO Network[J]. Computer Engineering, 2026, 52(5): 445-455.

https://www.ecice06.com/CN/Y2026/V52/I5/445

图/表 14

图1 DBE-YOLO网络结构

Fig.1 Network structure of DBE-YOLO

图2 BottleNeck及BottleNeck_CDEC结构

Fig.2 Structure of BottleNeck and BottleNeck_CDEC

图3 C2f及DBE-C2f模块结构

Fig.3 Structure of C2f and DBE-C2f module

图4 D-DADA算法流程

Fig.4 Procedure of D-DADA algorithm

图5 子策略增广示意图

Fig.5 Schematic diagram of sub-strategy augmentation

图6 检测效果热力图

Fig.6 Thermal map of detection effect

图7 不同光照条件下的检测效果

Fig.7 Detection effects under different lighting conditions

图8 不同拍摄角度下的检测效果

Fig.8 Detection effects under different shooting angles

图9 遮挡条件下的检测效果

Fig.9 Detection effects under occlusion conditions

参考文献 29

1	刘卫新, 胡春华. 电表异常状态监测方法研究. 自动化仪表, 2024, 45 (3): 64- 69.
	LIU W X , HU C H . Research on abnormal state monitoring method of electric energy metering device. Process Automation Instrumentation, 2024, 45 (3): 64- 69.
2	王以良, 周鹏, 叶卫, 等. 基于金字塔网络的非侵入式负荷辨识及其隐私保护方案. 计算机工程, 2024, 50 (5): 182- 189. doi: 10.19678/j.issn.1000-3428.0067381
	WANG Y L , ZHOU P , YE W , et al. Non-intrusive load monitoring and its privacy-preserving scheme based on pyramid network. Computer Engineering, 2024, 50 (5): 182- 189. doi: 10.19678/j.issn.1000-3428.0067381
3	李云鹏, 金旭荣, 张鑫瑞. 基于并行计算的电表故障诊断技术研究. 电测与仪表, 2022, 59 (12): 96- 102.
	LI Y P , JIN X R , ZHANG X R . Research on fault diagnosis technology of electric energy metering device based on parallel computing. Electrical Measurement & Instrumentation, 2022, 59 (12): 96- 102.
4	童光华, 李宁, 杨晨, 等. 面向非技术性损失的用电异常检测方法分析. 自动化与仪器仪表, 2019 (2): 122- 124.
	TONG G H , LI N , YANG C , et al. Analysis of electricity anomaly detection method for non technical loss. Automation & Instrumentation, 2019 (2): 122- 124.
5	杨劲业, 王昕, 郑益慧, 等. 基于ViBe检测的输电线路防外力破坏预警模型. 电气自动化, 2019, 41 (6): 36-39, 69.
	YANG J Y , WANG X , ZHENG Y H , et al. ViBe detection-based early warning model for prevention of external damage of transmission lines. Electrical Automation, 2019, 41 (6): 36-39, 69.
6	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proceedings of the European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 21-37.
7	REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1804.02767.
8	周自强, 陈强, 马必焕, 等. 一种改进的YOLO目标检测方法在电缆设备异常状态识别中的应用. 电测与仪表, 2020, 57 (2): 14- 20.
	ZHOU Z Q , CHEN Q , MA B H , et al. An improved YOLO target detection method with its application in cable device abnormal condition recognition. Electrical Measurement & Instrumentation, 2020, 57 (2): 14- 20.
9	陆煜锌, 赵云, 肖勇, 等. 面向数字孪生低压配电网的基于Gabor-YOLO算法的架空线高效识别方法研究. 电测与仪表, 2023, 60 (3): 40- 46.
	LU Y X , ZHAO Y , XIAO Y , et al. Research on efficient identification method of overhead lines based on Gabor-YOLO algorithm for digital twin low-voltage distribution network. Electrical Measurement & Instrumentation, 2023, 60 (3): 40- 46.
10	LUAN S Z , CHEN C , ZHANG B C , et al. Gabor convolutional networks. IEEE Transactions on Image Processing, 2018, 27 (9): 4357- 4366. doi: 10.1109/TIP.2018.2835143
11	杨露露, 马萍, 王聪, 等. 结合特征重用与重建的YOLO绝缘子检测方法. 计算机工程, 2024, 50 (7): 303- 313. doi: 10.19678/j.issn.1000-3428.0068244
	YANG L L , MA P , WANG C , et al. Insulator detection method using YOLO combining feature reuse and reconstruction. Computer Engineering, 2024, 50 (7): 303- 313. doi: 10.19678/j.issn.1000-3428.0068244
12	TAN M X, PANG R M, LE V L. Efficientdet: scalable and efficient object detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2020: 10781-10790.
13	吴庆, 胡明, 孟子健, 等. 电力应急通信非合作博弈资源分配方法. 电讯技术, 2024, 64 (2): 266- 272.
	WU Q , HU M , MENG Z J , et al. Power emergency communication resource allocation based on non-cooperative game. Telecommunication Engineering, 2024, 64 (2): 266- 272.
14	VARGHESE R M S. YOLOv8: a novel object detection algorithm with enhanced performance and robustness[C]//Proceedings of the 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems. Chennai, India: [s. n.], 2024: 1-6.
15	CONNOR S , KHOSHGOFTAAR TM . A survey on image data augmentation for deep learning. Journal of Big Data, 2019, 6 (1): 1- 48. doi: 10.1186/s40537-018-0162-3
16	LEMLEY J , BAZRAFKAN S , CORCORAN P . Smart augmentation learning an optimal data augmentation strategy. IEEE Access, 2017, 5, 5858- 5869. doi: 10.1109/ACCESS.2017.2696121
17	CUBUK E D, ZOPH B, MANE D, et al. Autoaugment: learning augmentation strategies from data[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2019: 113-123.
18	FAN L, FENG X, WANG H. An adaptive formulation-based many-objective evolutionary algorithm for multi-scenario optimization in data enrichment[C]//Proceedings of the 2021 IEEE Congress on Evolutionary Computation. Washington D. C., USA: IEEE Press, 2021: 1888-1895.
19	LIM S, KIM I, KIM T, et al. Fast AutoAugment[C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems. Washington D. C., USA: IEEE Press, 2019: 6662-6672.
20	LI Y, HU G, WANG Y. DADA: differentiable automatic data augmentation[C]//Proceedings of the European Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2020: 580-595.
21	LIN TY, MAIRE M, BELONGIE S, et al. Microsoft coco: Common objects in context[C]//Proceedings of the European Conference on Computer Vision. Berlin, Germany: Springer, 2014: 740-755.
22	ZHANG J, ZHENG K, ZHOU F, et al. Intelligent investigating technology for abnormal operation of wiring terminals of electric energy metering device based on deep learning[C]//Proceedings of the 2nd IEEE International Conference on Information Technology, Big Data and Artificial Intelligence. Washington D. C., USA: IEEE Press, 2021: 859-864.
23	张上, 王恒涛, 冉秀康. 基于YOLOv8的轻量化交通标志检测方法. 电子测量技术, 2022, 45 (8): 129- 135.
	ZHANG S , WANG H T , RAN X K . Lightweight traffic sign detection algorithm based on YOLOv5. Electronic Measurement Technology, 2022, 45 (8): 129- 135.
24	FANG W , WANG L , REN P . Tinier-YOLO: a real-time object detection method for constrained environments. IEEE Access, 2019, 8, 1935- 1944.
25	SELVARAJU RR, COGSWELL M, DAS A, et al. Grad-cam: visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2017, 618-626.
26	WANG C Y, BOCHKOVSKIY A, LIAO H Y. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2023: 7464-7475.
27	CARION N, MASSA F, SYNNAEVE G, et al. End-to-end object detection with transformers[C]//Proceedings of the European Conference on Computer Vision. Berlin, Germany: Springer, 2020: 553-562.
28	张英俊, 甘望阳, 谢斌红, 等. 融合多尺度特征与注意力的小样本目标检测. 小型微型计算机系统, 2025, 46 (3): 689- 696.
	ZHANG Y J , GAN W Y , XIE B H , et al. Few-shot object detection integrating multi-scale feature and attention. Journal of Chinese Computer Systems, 2025, 46 (3): 689- 696.
29	赵滨淋, 孙玲, 陈功, 等. SD-YOLO: 一种多尺度小目标检测算法. 电光与控制, 2025, 32 (7): 21- 26.
	ZHAO B L , SUN L , CHEN G , et al. SD-YOLO: a multi-scale small target detection algorithm. Electronics Optics & Control, 2025, 32 (7): 21- 26.

[1]	刘泽, 宋廷伦, 石先让, 苏洋, 赵群. 基于改进RT-DETR的路面异常检测技术研究[J]. 计算机工程, 2026, 52(4): 187-199.
[2]	董方和, 石琼, 师智斌. 基于集成学习与异常检测的对抗流量检测方法[J]. 计算机工程, 2026, 52(2): 275-286.
[3]	吴杰辉, 柳毅. 面向多元时间序列的联合优化异常检测模型[J]. 计算机工程, 2025, 51(9): 166-176.
[4]	徐式芃, 王雷, 盛捷. 基于知识图谱的异常个体提前识别模型研究[J]. 计算机工程, 2025, 51(9): 59-70.
[5]	王光明, 李冬青, 蒋从锋. 不平衡数据集下的数据中心网络流量异常检测[J]. 计算机工程, 2025, 51(8): 227-237.
[6]	陈梓延, 王晓龙, 何迪, 安国成. 基于改进YOLOv8的轻量化车辆检测网络[J]. 计算机工程, 2025, 51(5): 314-325.
[7]	张安勤, 丁志锋. 融合动态图嵌入和Transformer自编码器的网络异常检测[J]. 计算机工程, 2025, 51(4): 47-56.
[8]	牛渲文, 杜晔, 杨明松, 李昂, 黎妹红. 基于层次联邦与动态权重的卫星网络异常检测方法[J]. 计算机工程, 2025, 51(12): 210-220.
[9]	邸钦渤, 陈劭力, 时良仁. GRD: 基于GNN和扩散模型的多变量时序数据异常检测算法[J]. 计算机工程, 2025, 51(11): 35-44.
[10]	周炫余, 吴莲华, 郑勤华, 肖天星, 王紫璇, 张思敏. 联合语义提示和记忆增强的弱监督跳绳视频异常检测方法[J]. 计算机工程, 2024, 50(7): 87-95.
[11]	李永飞, 李铭洋, 常鑫, 曹可欣. 基于可解释性深度学习的物联网水质监测数据异常检测[J]. 计算机工程, 2024, 50(6): 179-187.
[12]	王俊, 赖会霞, 万玥, 张仕. 基于角度的图神经网络高维数据异常检测方法[J]. 计算机工程, 2024, 50(3): 156-165.
[13]	赵博超, 马嘉骏, 崔磊, 栾文鹏, 朱静. 基于改进VMD-XGBoost-BiLSTM组合模型的光伏发电异常检测[J]. 计算机工程, 2024, 50(3): 306-316.
[14]	宋航, 周凤, 熊伟. 基于自相关-变分对抗学习的物理系统异常检测[J]. 计算机工程, 2024, 50(12): 358-366.
[15]	陈仲磊, 伊鹏, 陈祥, 胡涛. 基于集成学习的系统调用实时异常检测框架[J]. 计算机工程, 2023, 49(6): 162-169,179.

选择文件类型/文献管理软件名称

选择包含的内容

基于D-DADA算法与DBE-YOLO网络的电表异常检测方法

Anomaly Detection Method for Electricity Meter Based on D-DADA Algorithm and DBE-YOLO Network

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 29

相关文章 15

编辑推荐

Metrics

本文评价

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

基于D-DADA算法与DBE-YOLO网络的电表异常检测方法

Anomaly Detection Method for Electricity Meter Based on D-DADA Algorithm and DBE-YOLO Network

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 29

相关文章 15

编辑推荐

Metrics

本文评价