湖南电力 ›› 2022, Vol. 42 ›› Issue (1): 17-22.doi: 10.3969/j.issn.1008-0198.2022.01.004

• 研究与试验 • 上一篇    下一篇

融合多级支持向量机的铁磁谐振和雷电过电压识别方法研究

赵洪彬1, 罗庆亮1, 李欣2, 向缨竹2, 何智强3, 邓化龙4   

  1. 1. 重庆大学电气工程学院,重庆 400044;
    2. 国网湖南省电力有限公司电力科学研究院,湖南 长沙 410007;
    3. 国网湖南省电力有限公司,湖南 长沙 410004;
    4. 湖南长高森源电力设备有限公司,湖南 衡阳 421200
  • 收稿日期:2021-09-02 修回日期:2021-09-15 发布日期:2025-08-05
  • 作者简介:赵洪彬(1989),男,硕士,工程师,主要从事电力系统过电压的研究。罗庆亮(1998),男,硕士,主要从事电力系统过电压的研究。
  • 基金资助:
    国网湖南省电力有限公司横向科研项目自然科学类(H20201372)

Research on Ferroresonance and Lightning Overvoltage Identification Method Based on Multi-Level Support Vector Machine

ZHAO Hongbin1, LUO Qingliang1, LI Xin2, XIANG Yingzhu2, HE Zhiqiang3, DENG Hualong4   

  1. 1. School of Electrical Engineering, Chongqing University, Chongqing 400044, China;
    2. State Grid Hunan Electric Power Company Limited Research Institute, Changsha 410007, China;
    3. State Grid Hunan Electric Power Company Limited, Changsha 410004, China;
    4. Hunan Changgao Senyuan Power Equipment Co., Ltd., Hengyang 421200, China
  • Received:2021-09-02 Revised:2021-09-15 Published:2025-08-05

摘要: 针对电力系统过电压经常发生的严重威胁设备及人身安全的问题,提出基于支持向量机的模式识别方法。以重庆市某110 kV变电站采集的过电压波形为基础,分析不同过电压波形的特征差异,进而采用时域分析和频域分析提取能够有效反映铁磁谐振和雷电过电压差异的6类特征量,最后将所提取的特征量应用到过电压的模式识别中。该识别方法采用的支持向量机理论基础简单、直观,仿真结果表明该方法能够正确识别铁磁谐振及雷电过电压,且具有识别速度快的优势。

关键词: 铁磁谐振过电压, 雷电过电压, 特征量提取, 支持向量机, 过电压识别

Abstract: Aiming at the frequent occurrence of overvoltage in power system, which seriously threatens the safety of equipment and personnel, a pattern recognition method based on support vector machine is proposed. Based on the overvoltage waveforms collected by Chongqing 110 kV substation, the characteristic differences of different overvoltage waveforms are analyzed, and then six kinds of characteristic quantities that can effectively reflect the difference between ferroresonance and lightning overvoltage are extracted by time domain analysis and frequency domain analysis. Finally, the extracted characteristic quantities are applied to overvoltage pattern recognition. The theoretical basis of support vector machine used in this recognition method is simple and intuitive. The simulation results show that this method can correctly identify ferroresonance and lightning overvoltage, and has the advantage of fast recognition speed.

Key words: ferroresonance overvoltage, lightning overvoltage, feature extraction, support vector machine, overvoltage identification

中图分类号: