基于SVM和SCADA大数据的母线负荷曲线识别方法
DOI:
作者:
作者单位:

1.长沙理工大学电气与信息工程学院;2.国网甘肃省电力公司;3.国网甘肃省电力公司电力科学研究院

作者简介:

通讯作者:

中图分类号:

TM863

基金项目:

国家自然科学基金(51977012);国家电网公司科技项目(52272218000X)


Bus Load Curve Recognition Method Based on SVM and SCADA Big Data
Author:
Affiliation:

1.School of Electrical and Informational Engineering,Changsha University of Science and Technology;2.State Grid Gansu Electric Power Company;3.State Grid Gansu Electric Power Corporation Research Institute

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对传统基于统计的负荷曲线分类准确性和费时实际,本文将非侵入式负荷监测与分解技术拓展应用于变电站母线负荷曲线分解。提出一种基于SVM和SCADA大数据的母线负荷曲线识别方法。首先分析典型行业负荷有功功率曲线变化过程,提取有功突变时间进行负荷预筛选,然后对有功功率波形进行傅里叶级数拟合,从而获取行业负荷特征标签,实现波形特征提取,最后采用支持向量机将变电站母线有功功率波形特征分类识别,实现行业负荷特征辨识。甘肃省电网某330kV变电站SCADA系统实际数据验证表明,该方法可有效获取母线负荷类别,从而提升负荷建模效率。

    Abstract:

    Aiming at the traditional statistics-based load curve classification accuracy and time-consuming reality, this paper extends non-intrusive load monitoring and decomposition technology to the substation bus load curve decomposition. A bus load curve recognition method based on SVM and SCADA big data is proposed. Firstly, analyze the change process of active power curve of typical industry load, extract active sudden change time for load pre-screening, then perform Fourier series fitting on the active power waveform to obtain industry load feature label, realize waveform feature extraction, and finally use support vector The machine will classify and identify the active power waveform characteristics of the substation bus to realize the identification of industry load characteristics. The actual data verification of the SCADA system of a 330kV substation in Gansu power grid shows that this method can effectively obtain the bus load category, thereby improving the efficiency of load modeling.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2020-11-30
  • 最后修改日期:2021-05-08
  • 录用日期:2021-05-10
  • 在线发布日期:
  • 出版日期: