Low-Voltage Substation Area Topology Recognition Method Based on AKNN Anomaly Detection and ADPC Clustering

doi:10.11930/j.issn.1004-9649.202307030

Abstract

Abstract:

The accurate record of topology information of the low-voltage station area is the basis for line loss analysis and three-phase imbalance control. Aiming at the problem of high cost and low efficiency of topology file investigation at present, a low-voltage substation area topology recognition method is proposed based on adaptive k nearest neighbor (AKNN) anomaly detection and adaptive density peaks clustering (ADPC). The similarity of voltage series between users in the low-voltage substation area is measured using dynamic time warping (DTW), and the abnormal relationship between users and transformer is checked and corrected with the AKNN anomaly detection algorithm. After getting the right relationship, the ADPC algorithm is used to identify the phase for users in the substation area. Finally, the case study of the actual substation area proves that the proposed method can effectively realize the topology identification of the low-voltage substation area without human parameter setting, and has high applicability and accuracy.

Key words: low-voltage substation area, user-transformer relationship, phase identification, adaptive k nearest neighbor, adaptive density peaks clustering

Ziyi SHI, Xiangyang XIA, Jiabin LIU, Yangyang GU, Yulong WANG, Jiayao HONG. Low-Voltage Substation Area Topology Recognition Method Based on AKNN Anomaly Detection and ADPC Clustering[J]. Electric Power, 2024, 57(5): 168-177.

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.electricpower.com.cn/EN/10.11930/j.issn.1004-9649.202307030

https://www.electricpower.com.cn/EN/Y2024/V57/I5/168

Figures/Tables 12

References 26

1	张丽强. 基于智能电表数据的供电网络拓扑识别方法研究[D]. 济南: 山东大学, 2020.
	ZHANG Liqiang. Research on topology identification method of power supply network based on smart meter data[D]. Jinan: Shandong University, 2020.
2	胡明月, 张慧芬, 苗淑平, 等. 基于改进鸽群算法的台区三相负荷不平衡相序调整方法[J]. 中国电力, 2022, 55 (9): 121- 128. DOI
	HU Mingyue, ZHANG Huifen, MIAO Shuping, et al. Phase sequence adjustment method for three-phase load imbalance in station area based on improved pigeon-inspired optimization[J]. Electric Power, 2022, 55 (9): 121- 128. DOI
3	李明维, 张传远, 马晓昆, 等. 低压配电台区网络拓扑和相别自动辨识技术研究[J]. 电气时代, 2018, (10): 62- 63.
4	张培玲, 赵可可. 基于单频通信的低压电力线通信系统设计与实现[J]. 中国电力, 2023, 56 (3): 118- 127, 136. DOI
	ZHANG Peiling, ZHAO Keke. Design and implementation of low voltage power communication system based on single frequency communication[J]. Electric Power, 2023, 56 (3): 118- 127, 136. DOI
5	单瑞卿, 盛阳, 苏盛, 等. 考虑攻击方身份的电力监控系统网络安全风险分析[J]. 电力科学与技术学报, 2022, 37 (5): 3- 16. DOI
	SHAN Ruiqing, SHENG Yang, SU Sheng, et al. Risk analysis of power system cyber security considering identity of malicious adversaries[J]. Journal of Electric Power Science and Technology, 2022, 37 (5): 3- 16. DOI
6	宋玮琼, 郭帅, 李冀, 等. 基于电压时序数据的配电台区户变关系智能识别[J]. 电力系统及其自动化学报, 2022, 34 (1): 48- 55.
	SONG Weiqiong, GUO Shuai, LI Ji, et al. Intelligent recognition of user-transformer relationship in distribution station area based on time series voltage data[J]. Proceedings of the CSU-EPSA, 2022, 34 (1): 48- 55.
7	连子宽, 姚力, 刘晟源, 等. 基于t-SNE降维和BIRCH聚类的单相用户相位及表箱辨识[J]. 电力系统自动化, 2020, 44 (8): 176- 184.
	LIAN Zikuan, YAO Li, LIU Shengyuan, et al. Phase and meter box identification for single-phase users based on t-SNE dimension reduction and BIRCH clustering[J]. Automation of Electric Power Systems, 2020, 44 (8): 176- 184.
8	陈招安, 黄纯, 张志丹, 等. 基于T型灰色关联度和KNN算法的低压配电网台区拓扑识别方法[J]. 电力系统保护与控制, 2021, 49 (1): 163- 169.
	CHEN Zhaoan, HUANG Chun, ZHANG Zhidan, et al. Topology identification method of a low voltage distribution network based on T-type grey correlation degree and KNN algorithm[J]. Power System Protection and Control, 2021, 49 (1): 163- 169.
9	崔雪原, 刘晟源, 金伟超, 等. 基于APAA和改进DBSCAN算法的户变关系及相位识别方法[J]. 电网技术, 2021, 45 (8): 3034- 3043.
	CUI Xueyuan, LIU Shengyuan, JIN Weichao, et al. Consumer-transformer relationship and phase identification based on APAA and improved DBSCAN algorithm[J]. Power System Technology, 2021, 45 (8): 3034- 3043.
10	徐明昕, 赵健, 王小宇, 等. 基于电压聚类和关联卷积的配电网户变关系识别方法[J]. 电力系统保护与控制, 2022, 50 (4): 92- 102. DOI
	XU Mingxin, ZHAO Jian, WANG Xiaoyu, et al. Transformer-customer identification method for a low-voltage distribution network based on voltage clustering and incidence convolution[J]. Power System Protection and Control, 2022, 50 (4): 92- 102. DOI
11	唐捷, 蔡永智, 李其霖, 等. 基于三相表特征约束聚类的低压台区用户相序识别方法[J]. 电力系统自动化, 2022, 46 (8): 60- 67.
	TANG Jie, CAI Yongzhi, LI Qilin, et al. Phase sequence identification method for users in low-voltage distribution-station area based on feature constraint clustering of three-phase meter[J]. Automation of Electric Power Systems, 2022, 46 (8): 60- 67.
12	ZHAO Y, TU Z D, XIAO Y, et al. Electrical principles of verification of the topological structure of a low-voltage distribution network based on voltage from an Advanced Metering Infrastructure[J]. The Journal of Engineering, 2019, 2019 (11): 8218- 8224. DOI
13	赖煊平. 基于数据驱动的低压配电网拓扑识别及应用研究[D]. 北京: 华北电力大学, 2022.
	LAI Xuanping. Data-driven topology identification and application of low-voltage distribution network[D]. Beijing: North China Electric Power University, 2022.
14	韩平平, 陈思远, 张楠, 等. 考虑分布式光伏的低压台区线损异常辨识方法[J]. 电力系统保护与控制, 2023, 51 (8): 140- 148. DOI
	HAN Pingping, CHEN Siyuan, ZHANG Nan, et al. Line loss anomaly identification method for low-voltage station area considering distributed PV[J]. Power System Protection and Control, 2023, 51 (8): 140- 148. DOI
15	李海林. 时间序列数据挖掘中的特征表示与相似性度量方法研究[D]. 大连: 大连理工大学, 2012.
	LI Hailin. Research on feature representation and similarity measure methods in time series data mining[D]. Dalian: Dalian University of Technology, 2012.
16	HAN T, PENG Q K, ZHU Z B, et al. A pattern representation of stock time series based on DTW[J]. Physica A:Statistical Mechanics and Its Applications, 2020, 550, 124161. DOI
17	王玲, 周南, 申鹏. 基于自适应k近邻的时间序列异常模式识别[J]. 计算机研究与发展, 2023, 60 (1): 125- 139.
	WANG Ling, ZHOU Nan, SHEN Peng. Time series anomaly pattern recognition based on adaptive k nearest neighbor[J]. Journal of Computer Research and Development, 2023, 60 (1): 125- 139.
18	WU X, KUMAR V. The top ten algorithms in data mining [M]. Florida: CRC Press, 2010: 151–158.
19	邹咸林. 自然最近邻居在高维数据结构学习中的应用[D]. 重庆: 重庆大学, 2011.
	ZOU Xianlin. Learning structure features in high dimensional data based on natural nearest neighbor[D]. Chongqing: Chongqing University, 2011.
20	PUKELSHEIM F. The three sigma rule[J]. The American Statistician, 1994, 48 (2): 88- 91.
21	卢鋆, 吴忠望, 王宇, 等. 基于kNN算法的异常行为检测方法研究[J]. 计算机工程, 2007, 33 (7): 133- 134, 138.
	LU Jun, WU Zhongwang, WANG Yu, et al. Research on abnormal behavior detection based on kNN algorithm[J]. Computer Engineering, 2007, 33 (7): 133- 134, 138.
22	RODRIGUEZ A, LAIO A. Clustering by fast search and find of density peaks[J]. Science, 2014, 344 (6191): 1492- 1496. DOI
23	QIAN P J. Fast graph-based relaxed clustering for large data sets using minimal enclosing ball[J]. IEEE Transactions on Systems, Man, and Cybernetics Part B, 2012, 42(3): 672–687.
24	肖勇, 赵云, 涂治东, 等. 基于改进的皮尔逊相关系数的低压配电网拓扑结构校验方法[J]. 电力系统保护与控制, 2019, 47 (11): 37- 43.
	XIAO Yong, ZHAO Yun, TU Zhidong, et al. Topology checking method for low voltage distribution network based on improved Pearson correlation coefficient[J]. Power System Protection and Control, 2019, 47 (11): 37- 43.
25	李亦非, 宋玮琼, 彭放, 等. 基于局部异常点检测算法的电能表飞走异常智能分析[J]. 电测与仪表, 2016, 53 (18): 69- 73.
	LI Yifei, SONG Weiqiong, PENG Fang, et al. The intelligent analysis on the trend anomaly of the electric energy meter based on LOF algorithm[J]. Electrical Measurement & Instrumentation, 2016, 53 (18): 69- 73.
26	廖志伟, 谢汛恺, 郑广昱, 等. 基于电压曲线形态特征聚类的低压台区相位辨识[J]. 电力系统自动化, 2023, 47 (7): 142- 149.
	LIAO Zhiwei, XIE Xunkai, ZHENG Guangyu, et al. Phase identification of low-voltage distribution station area based on morphological characteristic clustering of voltage curves[J]. Automation of Electric Power Systems, 2023, 47 (7): 142- 149.

户变关系真实情况	户变关系检验情况
户变关系真实情况	异常	正常
异常	$ {\lambda _{{\text{TP}}}} $	$ {\lambda _{{\text{FN}}}} $
正常	$ {\lambda _{{\text{FP}}}} $	$ {\lambda _{{\text{TN}}}} $

户变关系真实情况	户变关系检验情况
户变关系真实情况	异常	正常
异常	$ {\lambda _{{\text{TP}}}} $	$ {\lambda _{{\text{FN}}}} $
正常	$ {\lambda _{{\text{FP}}}} $	$ {\lambda _{{\text{TN}}}} $

用户编号	识别结果	正确率/%
1~6号、8~14号、16~25号、27~31号、33~37号	A相	100
38~40号、42~51号、53~61号、63~86号	B相	100
87~95号、97~113号、115~126号	C相	100

用户编号	识别结果	正确率/%
1~6号、8~14号、16~25号、27~31号、33~37号	A相	100
38~40号、42~51号、53~61号、63~86号	B相	100
87~95号、97~113号、115~126号	C相	100

方法	ARI
方法	正常情况	缺失率为5%	缺失率为10%	缺失率为15%
DTW	1.00	1.00	0.97	0.94
Pearson相似性（最近邻点插值）	1.00	0.95	0.87	0.82
Pearson相似性（线性插值）	1.00	0.97	0.92	0.85
欧式距离（最近邻点插值）	0.93	0.89	0.81	0.73
欧式距离（线性插值）	0.97	0.93	0.83	0.78