基于Q学习的电网运行断面动态生成方法
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TM743

基金项目:

广东省自然科学基金(2018A0303130134);中国南方电网有限责任公司科技项目(0000002019030101XT00035)


A novel dynamic detection approach for power system operation section based on Qlearning algorithm
Author:
Affiliation:

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:

    电网运行断面是电力系统运行控制的重要手段。面对当前繁多的电网运行断面智能生成方法,如何合理的选择已成为电网运行断面在线生成算法领域研究的重要内容。在此背景下,提出一种基于 Q 学习的电网运行断面动态生成方法。该方法的主要特征在于训练得到 Q 学习智能体,根据电网运行特征动态选择电网运行断面生成方法,以便充分利用不同生成方法在不同场景下的算法优势。最后,基于某电网数据构造的算例表明,动态生成方法能够通过优化选择不同场景下的生成算法,提升生成结果的准确率。对于应用样本集,该方法提高准确率近5.2%。

    Abstract:

    Power grid operation section is an important measure in power system operation control. Faced with the numerous intelligent generation methods of grid operation sections at present, how to make a reasonable choice has become an important content of the research in the field of online generation algorithms for grid operation sections. To solve this problem, a dynamic detection method for power grid operation section based on Qlearning algorithm is proposed. The main feature of this method is that the Qlearning agent is trained, and the grid operation section generation method is dynamically selected according to the grid operation characteristics, so as to make full use of the algorithm advantages of different generation methods in different scenarios. Finally, a case study based on the actual data in a certain provincial power grid shows that the dynamic detection method can improve the accuracy of the generated results by optimizing the selection of the detection algorithms in different scenarios. For the applied sample set, the method could improve the accuracy by nearly 5.2%.

    参考文献
    相似文献
    引证文献
引用本文

李豹,吴云亮,邓韦斯,李鹏.基于Q学习的电网运行断面动态生成方法[J].电力科学与技术学报,2021,(4):116-123. Li Bao, Wu Yunliang, Deng Weisi, Li Peng. A novel dynamic detection approach for power system operation section based on Qlearning algorithm[J]. JOURNAL OF EIECTRIC POWER SCIENCE AND TECHNOLOGY,2021,(4):116-123.

复制
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:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2021-08-28
  • 出版日期: