电器与能效管理技术 ›› 2025, Vol. 0 ›› Issue (1): 23-27.doi: 10.16628/j.cnki.2095-8188.2025.01.004

• 研究与分析 • 上一篇    下一篇

基于多传感信息融合的开关柜局放类型诊断研究

李刚毅   

  1. 中国铝业股份有限公司 广西分公司, 广西 百色 531400
  • 收稿日期:2024-09-18 出版日期:2025-01-30 发布日期:2025-03-17
  • 作者简介:李刚毅(1982—),男,工程师,主要从事氧化铝行业电气自动化管理和技术等方面的研究工作。

Research on Partial Discharge Type Diagnosis of Switchgear Based on Multi-Sensor Information Fusion

LI Gangyi   

  1. Guangxi Branch,China Aluminum Co.,Ltd., Baise 531400, China
  • Received:2024-09-18 Online:2025-01-30 Published:2025-03-17

摘要:

在开关柜运行过程中,绝缘缺陷可能导致局部放电现象的发生,如果不能及时发现并采取处理措施,就会进一步造成绝缘击穿故障。为了有效诊断局部放电类型,提出一种基于D-S证据理论的多传感信息融合技术,结合反向传播(BP)神经网络进行局部放电类型识别的方案。实验表明,经过多传感信息融合后放电类型识别准确率较单传感器模式平均提高约13%,整体识别准确率达到96.08%,验证所提算法的有效性及合理性。

关键词: 局部放电, 开关柜, 模式识别, 多传感信息融合

Abstract:

During the operation of the switchgear,the insulation defects may lead to the occurrence of partial discharge.If it can not be found and dealt with measures in time,it will further cause the insulation breakdown failure.In order to effectively diagnose the type of partial discharge,based on the multi-sensor information fusion technology of D-S evidence theory,a scheme combined with BP neural network for the partial discharge pattern recognition is proposed.Experiments show that the accuracy of discharge type recognition with the multi-sensor information fusion is about 13 percentage higher than that of single sensor mode.The overall recognition accuracy is better than 96.08%, which verifies the effectiveness and rationality of the algorithm.

Key words: partial discharge, switchgear, pattern recognition, multi-sensor information fusion

中图分类号: