基于孤立森林算法与大数据挖掘的配电网故障距离估计方法*

doi:10.11985/2022.01.023

电气工程学报 ›› 2022, Vol. 17 ›› Issue (1): 179-185.doi: 10.11985/2022.01.023

扫码分享

基于孤立森林算法与大数据挖掘的配电网故障距离估计方法^*

向绍斌(), 涂水员()

广西防城港核电有限公司电仪部防城港 538000

收稿日期:2021-06-01 修回日期:2021-09-30 出版日期:2022-03-25 发布日期:2022-05-06
作者简介:向绍斌,男,1976年生,高级工程师。主要研究方向为电力系统继电保护、发电机励磁、配电、高电压技术等。E-mail： xiangshaobin747899@163.com;
涂水员,男,1979年生,高级工程师。主要研究方向为电力系统继电保护、发电机励磁、电力系统电气自动化等。E-mail： tu00987690@126.com
基金资助:
*国家重点研发计划资助项目(2018YFB0904700)

Distribution Network Fault Distance Estimation Method Based On Isolated Forest Algorithm and Big Data Mining

XIANG Shaobin(), TU Shuiyuan()

Electrical Instrument Control Department, Guangxi Fangchenggang Nuclear Power Co., Ltd., Fangchenggang 538000

Received:2021-06-01 Revised:2021-09-30 Online:2022-03-25 Published:2022-05-06

摘要/Abstract

摘要：

针对配电网线路径向分布、分支线复杂特性,存在多重估计的问题,提出一种基于孤立森林算法与大数据挖掘的配电网故障距离估计方法。首先,构建“云-管-边-端”新型配电物联网架构,实现数据的高效管理。然后,在配电边缘代理装置中引入孤立森林算法提取故障时电压电流异常特征数据。最后,基于配电云主站汇集并获取海量装置中提取的异常特征数据集,利用基于代码库的大数据挖掘技术(Big data mining base on code repositories,BDMBCR)搜索故障事件间的关联关系,根据数据相似性归类分支分组,估算故障发生的具体位置。采用IEEE 34测试仿真系统进行试验。结果表明,提出的方法能够挖掘复杂数据特性,两相和三相接地故障多重估计长度相应减少了24%~27%和16%~20%,单相接地的故障影响距离仅为1.2 km,优于其他对比方法。

关键词: 配电物联网, 边缘计算, 孤立森林算法, 故障距离估计, 大数据挖掘

Abstract:

In order to solve the problem of multiple estimation due to the complex characteristics of distribution network, a fault distance estimation method based on isolated forest algorithm and big data mining is proposed. Firstly, a new distribution Internet of things architecture of “cloud management side end” is constructed to realize efficient data management. Then, the isolated forest algorithm is introduced into the distribution edge agent to extract the abnormal voltage and current characteristic data. Finally, based on the distribution cloud master station to collect and obtain the abnormal feature data set extracted from a large number of devices, using the big data mining technology based on code base (BDMBCR) to search the association between fault events, classify the branches and groups according to the data similarity, and estimate the specific location of the fault. The IEEE 34 test simulation system is used for the experiment. The results show that the proposed method can mine the complex data characteristics. The multiple estimation length of two-phase and three-phase grounding fault is reduced by 24%-27% and 16%-20% respectively. The single-phase grounding fault influence distance is only 1.2 km, which is better than other comparison methods.

Key words: Distribution power Internet of Things, edge computing, isolated forest algorithm, fault distance estimation, big data mining

中图分类号:

TM72

向绍斌, 涂水员. 基于孤立森林算法与大数据挖掘的配电网故障距离估计方法^*[J]. 电气工程学报, 2022, 17(1): 179-185.

XIANG Shaobin, TU Shuiyuan. Distribution Network Fault Distance Estimation Method Based On Isolated Forest Algorithm and Big Data Mining[J]. Journal of Electrical Engineering, 2022, 17(1): 179-185.

图/表 9

图1

图2

图3

图4

图5

图6

表1

图7

参考文献 16

[1]	李燕, 胡志坚, 仉梦林, 等. 基于不确定网络理论计及经济性与可靠性的主动配电网规划[J]. 电力系统自动化, 2019, 43(16):68-82.
	LI Yan, HU Zhijian, ZHANG Menglin, et al. Active distribution network planning considering economy and reliability based on uncertain network theory[J]. Automation of Electric Power Systems, 2019, 43(16):68-82.
[2]	郭晓斌, 许悦, 程乐峰, 等. 基于拓扑相似性分析的配电网网架可靠性评估[J]. 高电压技术, 2018, 44(11):3760-3769.
	GUO Xiaobin, XU Yue, CHENG Lefeng, et al. Reliability evaluation of distribution network based on topological similarity analysis[J]. High Voltage Engineering, 2018, 44(11):3760-3769.
[3]	PAU Marco, PATTI Edoardo, BARBIERATO Luca, et al. A cloud-based smart metering infrastructure for distribution grid services and automation[J]. Sustainable Energy,Grids and Networks, 2018, 15:14-25. doi: 10.1016/j.segan.2017.08.001
[4]	杜博. 云计算环境下的智能电网光通信网络安全的研究[J]. 自动化与仪器仪表, 2018, 19(1):22-26.
	DU Bo. Research on smart grid optical communication network security in cloud computing environment[J]. Automation and Instrumentation, 2018, 19(1):22-26.
[5]	FARIAS P E, DE MORAIS A P, ROSSINI J P, et al. Non-linear high impedance fault distance estimation in power distribution systems:A continually online-trained neural network approach[J]. Electric Power Systems Research, 2018, 157:20-28. doi: 10.1016/j.epsr.2017.11.018
[6]	ASGHARIGOVAR S, POURGHASEM P, SEYEDI H. High impedance fault protection scheme for smart grids based on WPT and ELM considering evolving and cross-country faults[J]. International Journal of Electrical Power & Energy Systems, 2019, 107:412-421. doi: 10.1016/j.ijepes.2018.12.019
[7]	薛士敏, 陆俊弛, 刘冲, 等. 基于虚拟线路阻抗的MMC-HVDC输电系统单端故障测距方法[J]. 电网技术, 2019, 43(8):2868-2875.
	XUE Shimin, LU Junchi, LIU Chong, et al. Single terminal fault location method of MMC-HVDC transmission system based on virtual line impedance[J]. Power System Technology, 2019, 43(8):2868-2875.
[8]	ZHOU Hui, WANG Zhong, SHI Chunqing, et al. Study of fault diagnosis distribution network based on rough set and artificial intelligence[J]. Journal of Physics:Conference Series, 2021, 1754(1):12-16.
[9]	谢李为, 李勇, 罗隆福, 等. 基于距离矩阵与分支系数的配电网故障定位方法[J]. 中国电机工程学报, 2020, 40(7):2180-2191,2397.
	XIE Liwei, LI Yong, LUO Longfu, et al. Distribution network fault location method based on distance matrix and branch coefficient[J]. Proceedings of the CSEE, 2020, 40(7):2180-2191,2397.
[10]	ZHAO Shutao, YAN Xiao, WANG Bo, et al. Research on reliability evaluation method of DC circuit breaker based on Markov model[J]. Electric Power Systems Research, 2019, 173(8):1-5. doi: 10.1016/j.epsr.2019.04.005
[11]	余翔, 陈国洪, 李霆, 等. 基于孤立森林算法的用电数据异常检测研究[J]. 信息技术, 2018, 42(12):96-100.
	YU Xiang, CHEN Guohong, LI Ting, et al. Research on anomaly detection of power consumption data based on isolated forest algorithm[J]. Information Technology, 2018, 42(12):96-100.
[12]	FARUGHIAN A, KUMPULAINEN L, KAUHANIEMI K. Review of methodologies for earth fault indication and location in compensated and unearthed MV distribution networks[J]. Electric Power Systems Research, 2018, 154:373-380. doi: 10.1016/j.epsr.2017.09.006
[13]	SWETAPADMA A, YADAV A. A hybrid method for fault location estimation in a fixed series compensated lines[J]. Measurement, 2018, 123:8-18. doi: 10.1016/j.measurement.2018.03.029
[14]	薛士敏, 齐金龙, 刘冲, 等. 直流微网接地方式及新型保护原理[J]. 电网技术, 2018, 42(1):48-55.
	XUE Shimin, QI Jinlong, LIU Chong, et al. DC microgrid grounding mode and new protection principle[J]. Power System Technology, 2018, 42(1):48-55.
[15]	LI Mengshi, YU Da, CHEN Ziming, et al. A data-driven residual-based method for fault diagnosis and isolation in wind turbines[J]. IEEE Transactions on Sustainable Energy, 2018, 10(2):895-904. doi: 10.1109/TSTE.2018.2853990
[16]	BEHESHTAEIN S, CUZNER R, SAVAGHEBI M, et al. Fault location in microgrids:A communication-based high-frequency impedance approach[J]. IET Generation,Transmission & Distribution, 2018, 13(8):1229-1237. doi: 10.1049/iet-gtd.2018.5166

故障类型	常规故障定位	案例 1		案例2
故障类型	长度h/km	长度h/km	减少比重d(%)	长度h/km	减少比重d(%)
A-G	44.673	3.57	92.01	1.274	97.15
B-G	30.085	3.9	87.04	0.97	96.78
C-G	3.57	2.86	19.89	0.99	72.27
AB	3.57	3.01	15.69	1.97	44.82
BC	3.57	3.26	8.68	1.85	48.18
CA	3.57	3.09	13.45	1.92	46.22
AB-G	3.57	3.57	0	2.70	24.37
BC-G	3.57	3.57	0	2.71	24.09
CA-G	3.57	3.57	0	2.63	26.33
ABC	3.57	3.57	0	2.87	19.61
ABC-G	3.57	3.57	0	2.98	16.53

基于孤立森林算法与大数据挖掘的配电网故障距离估计方法^*

Distribution Network Fault Distance Estimation Method Based On Isolated Forest Algorithm and Big Data Mining

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 16

相关文章 3

编辑推荐

Metrics

本文评价

[1]	李兴, 侯伟, 杨宇, 王鹏, 苏海军, 付铭明. 论电力基建业务中边缘计算的应用与实践[J]. 电气工程学报, 2020, 15(4): 153-160.
[2]	田鹏,秦贺,侯乃乾,刘志强. 基于边缘计算和广域同步测量技术的配电网故障定位方法 ^*[J]. 电气工程学报, 2020, 15(1): 83-88.
[3]	巫江虹, 刘超鹏, 梁志豪, 张才俊. 房间空调器长效运行性能预测及优化方案的研究[J]. 机械工程学报, 2015, 51(18): 158-166.

1：For N个配电边缘代理装置 do
2：设定树高的范围$j=ceiling(\text{lb}({{y}_{n}}))$
3： For 配电边缘代理装置的$\left\{ {{x}_{n}} \right\}$棵树 do
4：当前树高i=0
5： If i >j或只有一个数据 then
6： Return 叶子节点
7： Else
8：任意选取一个属性d,并且在其中任意选取一个划分数值f。
9：根据与f大小的对比结果排列子节点,如果样本数值不超过f,则放置于此节点的左子节点,剩下的放置于右子节点。
10： i=i+1
11：重复上述步骤,以得到iTree。
12： End if
13：将得到的iTree融入孤立森林集合。
14： End for
15：End for