Scholarlink

<p>Partial discharge faults in switchgear pose significant challenges to power maintenance due to late detection, difficult information collection, and inaccurate localization. To address the monitoring of switchgear partial discharge, this study proposes a real-time monitoring model based on a hybrid spatio-temporal convolutional network, integrating deep learning algorithms. The model accurately identifies minute discharge currents and locates faults by analyzing pyroelectric particle concentration. Experimental results show an average localization accuracy of 97.35% across 64 test points, a positive correlation between discharge charge quantity and particle concentration, and an overall positioning error of less than 3. This research provides a reliable technical basis for intelligent partial discharge monitoring in switchgear and points to new directions for the development of power equipment condition diagnosis and early warning systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on deep learning-based multi-parameter integrated monitoring technology of switchgear insulation pyroelectric particle partial discharge

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月17日

查看全文
  • Song Wu,
  • Zongxi Xie,
  • Ting Yang,
  • Na Zhan,
  • Dengwei Yu

摘要

Partial discharge faults in switchgear pose significant challenges to power maintenance due to late detection, difficult information collection, and inaccurate localization. To address the monitoring of switchgear partial discharge, this study proposes a real-time monitoring model based on a hybrid spatio-temporal convolutional network, integrating deep learning algorithms. The model accurately identifies minute discharge currents and locates faults by analyzing pyroelectric particle concentration. Experimental results show an average localization accuracy of 97.35% across 64 test points, a positive correlation between discharge charge quantity and particle concentration, and an overall positioning error of less than 3. This research provides a reliable technical basis for intelligent partial discharge monitoring in switchgear and points to new directions for the development of power equipment condition diagnosis and early warning systems.