作者：侯思祖¹, 徐岩¹, 李柏奎², 郝淑敏¹

作者单位：1. 华北电力大学电气与电子工程学院，河北 保定 071003;
2. 中国电力科学研究院有限公司，北京 100192

关键词：故障选线;极坐标变换;图像融合;YOLO网络

摘要：

针对现有的配电网故障选线困难的问题，提出一种基于YOLO网络的配电网故障选线方法。首先，使用三相电流构建极坐标二维图像，充分提取特征信息；之后，对各线路三相电流的极坐标图像进行像素级融合，并以不同颜色区分，在进一步加强图像特征的同时，降低原始图像的冗余度；最后，使用YOLO神经网络对融合图像进行特征提取，训练得到最优的模型文件，利用该模型实现故障选线。将该方法与现有的故障选线结果进行对比，结果表明，该方法选线准确率可以达到99.95%，选线时间12.9 ms，明显优于其他故障选线方案，且该方案不受故障时刻、故障类型和过渡电阻等因素的影响，可满足配电网故障选线的准确度和可靠性需求。

YOLO based fault line selection method for distribution network
HOU Sizu¹, XU Yan¹, LI Baikui², HAO Shumin¹
1. School of Electrical and Electronic Engineering, North China Electric Power University, Baoding 071003, China;
2. China Electric Power Research Institute, Beijing 100192, China
Abstract: In order to make full use of the advantages of deep learning for distribution network fault line selection, a distribution network fault line selection method based on YOLO network is proposed. Firstly, the polar two-dimensional image is constructed by using three-phase current, and the feature information is fully extracted. Then, the polar images of three-phase current of each feeder are fused at pixel level and distinguished by different colors, which further strengthens the image features and reduces the redundancy of the original image. Finally, YOLO neural network is used to extract the features of the fused image, train the optimal model file, and use the model to realize fault line selection. Comparing the proposed method with the existing fault line selection results, the results show that the accuracy of the proposed method can reach 99.95% and the line selection time is only about 12.9 ms, which is obviously better than other fault line selection schemes. Moreover, the scheme is not affected by factors such as fault time, fault type and transition resistance, and can meet the accuracy and reliability requirements of fault line selection in distribution network.
Keywords: fault line selection;polar coordinate transformation;image fusion;YOLO network
2024, 50(2):117-125  收稿日期: 2022-03-08;收到修改稿日期: 2022-06-01
基金项目: 国家重点研发计划(2018YFF01011900)
作者简介: 侯思祖(1962-),男,山西运城市人,教授,博士生导师,研究方向为配电网及其主要设备故障诊断等。
参考文献
[1] 马向辉, 马亚林, 李兆众. 小电流接地系统单相接地故障与选线分析[J]. 集成电路应用, 2020, 37(12): 86-87
MA X H, MA Y L, LI Z Z. Analysis of single phase grounding fault and line selection in small current grounding system[J]. Applications of IC, 2020, 37(12): 86-87
[2] 薛颢, 李延军, 程虎. 一种利用零序功率方向的小电流接地系统单相接地故障选线判据[J]. 广东电力, 2018, 31(4): 132-138
XUE H, LI Y J, CHENG H. Criterion for single-phase grounding fault line selection for small current grounding system by using zero sequence power direction[J]. Guangdong Electric Power, 2018, 31(4): 132-138
[3] 郭威. 小电流接地系统五次谐波选线法研究[D]. 陕西: 西安科技大学, 2012.
GUO W. Fault line selection for ineffectively earthed distribution systems based on the fifth harmonic current[D]. Shaanxi: Xi’an University of Science and Technology, 2012.
[4] 龙茹悦, 黄纯, 汤涛, 等. 一种谐振接地系统的配电线路接地故障选线新方法[J]. 电力系统保护与控制, 2019, 47(21): 21-29
LONG R Y, HUANG C, TANG T, et al. Line detection method for grounding fault in resonant grounding systems[J]. Power System Protection and Control, 2019, 47(21): 21-29
[5] 张春秋, 杨洪耕. 考虑三相不平衡电流影响的改进导纳法故障选线[J]. 四川电力技术, 2016, 9(3): 63-66
ZHANG C Q, YANG H G. Fault line selection with improved admittance considering the impact of three- phase unbalance current[J]. Sichuan Electric Power Technology, 2016, 9(3): 63-66
[6] 王建元, 张峥. 基于注入信号与小波能量的小电流接地故障选线研究[J]. 电测与仪表, 2018, 55(5): 28-32
WANG J Y, ZHANG Z. Research of fault line selection in small current grounding system based on injected signal and wavelet energy[J]. Electrical Measurement & Instrumentation, 2018, 55(5): 28-32
[7] YANG X, LIU W. A new line selection method based on intrinsic mode energy of phase current change[J]. Electric Power Science and Technology, 2018, 33(4): 149-154
[8] 杨婧, 辛明勇, 欧家祥, 等. 基于大数据的配电网线损定位与评估方法研究[J]. 中国测试, 2019, 45(7): 19-24
YANG J, XIN Y M, OU J X, et al. Distribution network line loss location and evaluation method study based on big data[J]. China Measurement and Test, 2019, 45(7): 19-24
[9] 丁楠楠, 胡旭晓, 吴跃成, 等. 基于深度学习的输电线路外力破坏检测方法研究[J]. 软件工程, 2022, 25(1): 14-17
DING N N, HU X X, WU Y C, et al. Research on external force damage detection method of transmission line based on deep learning[J]. Software Engineer, 2022, 25(1): 14-17
[10] 孟安波, 葛佳菲, 李德强, 等. 基于纵横交叉算法的神经网络配电网故障选线研究[J]. 电力系统保护与控制, 2016, 44(21): 694-700
MENG A B, GE J F, LI D Q, et al. Research on fault line selection of distribution network using RBF neural network based on crisscross optimization algorithm optimization[J]. Power System Protection and Control, 2016, 44(21): 694-700
[11] 严凤, 李双双. 基于 C 型行波与 SVM 的配电线路故障定位[J]. 电力系统及其自动化学报, 2016, 28(1): 86-90
YAN F, LI S S. Composite fault location method based on C-traveling wave and SVM for distribution lines[J]. Proceedings of the CSU-EPSA, 2016, 28(1): 86-90
[12] 吉兴全, 陈金硕, 张玉敏, 等. 基于CNN-SVM的配电网故障分类研究[J]. 智慧电力, 2022, 50(1): 94-100
JI Q X, CHEN J S, ZHANG Y M, et al. Fault Classification in Distribution Network Based on CNN-SVM[J]. Smart Power, 2022, 50(1): 94-100
[13] 程文傲, 徐明, 高金峰. 小电流接地系统单相接地故障选线空间域图像生成及融合方法[J]. 电力自动化设备, 2021, 41(7): 97-103
CHENG W A, XU M, GAO J F. Spatial domain image generation and fusion method of single-phase grounding fault line selection for small current grounding system[J]. Electric Power Automation Equipment, 2021, 41(7): 97-103
[14] YAN B, FAN P, LEI X, et al. A real-time apple targets detection method for picking robot based on improved YOLOv5[J]. Remote Sensing, 2021, 13(9): 1619-1619
[15] 杨仁桓, 林中森, 夏良标, 等. 一种不停运配电网接地故障定位仪设计[J]. 中国测试, 2015, 203(6): 64-67
YANG R H, LIN Z S, XIA L B, et al. Design of grounding-fault locators for not-outage power distribution grid[J]. China Measurement and Test, 2015, 203(6): 64-67