作者：李天楚1,2, 伍智鹏1,2, 刘红岩1,2
作者单位：1. 海南电网有限责任公司电力科学研究院，海南 海口 570311;
2. 海南省电网理化分析重点实验室，海南 海口 570311
关键词：电压暂降;数据冗余;同源识别;Hausdorff距离;多维尺度分析算法;OPTICS聚类
摘要：
为合理评估、治理电压暂降问题，电网公司大量安装电能质量监测终端，实现电压暂降监测和感知，但同时造成暂降事件被不同监测终端重复记录的数据冗余问题。为此，该文提出基于Hausdorff距离的电压暂降波形相似度计算方法。将波形相似度和持续时间作为同源识别特征，构建电压暂降同源识别特征矩阵；提出基于多维尺度分析（multi-dimensional scaling, MDS）算法的同源识别矩阵数据降维方法，保证识别准确度，提高计算效率。提出基于OPTICS聚类算法的电压暂降同源识别方法，通过聚类可达图得到同源识别结果。最后，选取我国中部某地区电网的实测数据验证所提方法的可行性与准确性。
Same-source identification method of voltage sags based on Hausdorff distance and OPTICS clustering
LI Tianchu1,2, WU Zhipeng1,2, LIU Hongyan1,2
1. Electric Power Research Institute of Hainan Power Grid Co., Ltd., Haikou 570311, China;
2. Key Laboratory of Physical and Chemical Analysis for Electric Power of Hainan Province, Haikou 570311, China
Abstract: In order to evaluate and manage the voltage sag problem reasonably, the power grid company installed a large number of power quality monitoring terminals to realize the voltage sag monitoring and perception, but at the same time it caused the data redundancy problem that the sag events were repeatedly recorded by different monitoring terminals. The similarity calculation method of voltage sags’ waveforms based on Hausdorff distance is proposed. The waveform similarity and duration are used as the same-source identification characteristics to construct the voltage sags’ same-source identification matrix, and the same-source identification matrix data reduction method based on the multi-dimensional scaling algorithm is proposed to ensure the identification accuracy and improve the calculation efficiency. A same-source identification method of voltage sags based on OPTICS clustering algorithm is proposed, and the same-source identification result is obtained through clustering reachability-plot. Finally, the measurement data of the power grid of a region in the middle of China are selected to verify the feasibility and correctness of the proposed method.
Keywords: voltage sag;data redundancy;same-source identification;Hausdorff distance;multi-dimensional scaling algorithm;OPTICS clustering
2022, 48(4):110-116,172 收稿日期: 2021-03-05;收到修改稿日期: 2021-05-17
基金项目: 海南电网科技项目（073000KK52190007）
作者简介: 李天楚（1988-），女，河北唐山市人，高级工程师，硕士，研究方向为电能质量、电磁环境与仪器仪表检测技术
参考文献
[1] 沈煜, 邬捷龙, 王斌, 等. 影响电压暂降问题的多源属性排序方法[J]. 中国测试, 2020, 46(2): 16-22
[2] 陶顺, 肖湘宁. 电压暂降对配电系统可靠性影响及其评估指标的研究[J]. 中国电机工程学报, 2005, 25(21): 63-69
[3] 张逸, 杨洪耕, 叶茂清. 基于分布式文件系统的海量电能质量监测数据管理方案[J]. 电力系统自动化, 2017, 38(2): 102-108
[4] 肖先勇, 桂良宇, 李成鑫, 等. 基于Wasserstein距离的多电压暂降事件同源检测方法[J]. 电网技术, 2020, 44(12): 4684-4693
[5] 汪颖, 谢佳妮, 邓凌峰, 等. 基于典型波形特征与改进DBSCAN的电压暂降同源识别方法[J/OL]. 电力系统自动化. https://kns.cnki.net/kcms/detail/32.1180.TP.20210204.1716.006.html.
[6] AUNG M T, MILANOVIC J V. The influence of transformer winding connections on the propagation of voltage sags[J]. IEEE Transactions on Power Delivery, 2006, 21(1): 262-269
[7] BLAIR S M, BOOTH C D, WILLIAMSON G, et al. Automatically detecting and correcting errors in power quality monitoring data[J]. IEEETransactions on Power Delivery, 2017, 32(2): 1005-1013
[8] 陈众励, 许维胜. 电压中断与电压暂降的成因及其防治[J]. 电工技术学报, 2015, 30(1): 518-510
[9] Bollen M H J. 理解电能质量问题: 电压暂降与短时中断[M]. 肖先勇, 汪颖, 译. 北京: 中国电力出版社, 2016.116-139.
[10] IEEE guide for voltage sag indices: IEEE 1564-2014[S], 2014.
[11] 汪颖, 周杨, 肖先勇, 等. 电压暂降问题研究现状及面临的挑战[J]. 供用电, 2018, 35(2): 2-9
[12] BASTOS A F, LAO K W, TODESCHINI G, et al. Accurate identification of point-on-wave inceptionand recovery instants of voltage sags and swells[J]. IEEE Transactions on Power Delivery, 2019, 34(2): 551-560
[13] 翁汉琍, 刘华, 林湘宁, 等. 基于Hausdorff距离算法的变压器差动保护新判据[J]. 中国电机工程学报, 2018, 38(2): 475-484
[14] 王家驹, 万忠兵. 何仲潇, 等. 基于多维尺度分析和改进K-means的台户关系辨识方法[J]. 电力系统及其自动化, 2020, 42(2): 56-59
[15] 刘振焘, 徐建平, 吴敏, 等. 语音情感特征提取及其降维方法综述[J]. 计算机学报, 2018, 41(12): 2833-2851
[16] 路郁. 聚类算法在雷电预警及雷灾分析中的应用[D]. 南京: 南京信息工程大学, 2014.
[17] 党秋月, 陆月明. 基于OPTICS可达图的自动识别簇方法[J]. 计算机应用, 2012, 32(S2): 19-21,47
