电器与能效管理技术 ›› 2020, Vol. 593 ›› Issue (8): 68-74.doi: 10.16628/j.cnki.2095-8188.2020.08.013

• 配网技术与系统 • 上一篇    下一篇

基于DBN的配电网高阻接地故障识别

刘炳南   

  1. 国网福建省电力有限公司 长乐供电公司,福建 福州 350202
  • 收稿日期:2020-02-20 出版日期:2020-08-30 发布日期:2020-09-10
  • 作者简介:刘炳南(1993—),男,主要从事配电网及其自动化技术研究。

High Impedance Fault Identification in Distribution Network Based on Deep Belief Network

LIU Bingnan   

  1. Changle Electric Power Supply Company,State Grid Fujian Electric Company,Fuzhou 350202,China
  • Received:2020-02-20 Online:2020-08-30 Published:2020-09-10

摘要:

配电网运行时受现实条件的影响,接触到电阻大的物质从而发生高阻接地故障(HIF)时,故障电流微弱,波形特征不明显,常规保护装置无法准确切除故障。利用希尔伯特-黄变换(HHT)带通滤波对母线三相电压和零序电压进行处理,构造分块时频能量谱,输入到深度置信网络进行训练和分类识别。最后,通过搭建的仿真模型,验证算法具有较高的准确率。

关键词: 配电网, 高阻接地故障, HHT带通滤波, 深度置信网络

Abstract:

With the development of distribution network,its scale is getting larger and larger,and its structure has become complex and diversified.Therefore,more attention has been paid to the safety,reliability and economic operation of distribution network.The distribution network operation is affected by the real conditions,and high impedance faults occur when contact with high impedance material.When the fault occurs,the fault current is weak and the waveform characteristics are not obvious,so the conventional protection device can't accurately remove the fault.The fault is often accompanied by harmonic and arc phenomena,which will lead to new fault points and further expand the accident if it exists for a long time.In this paper,the HHT band-pass filter is used to process the three-phase voltage and zero sequence voltage of the bus,and to construct the time-frequency energy spectrum of block.After processing,it is input into deep belief network for training and classification.The simulation model established in this paper proves that this algorithm has high accuracy.

Key words: distribution network, high impedance fault(HIF), HHT band-pass filter, deep belief network

中图分类号: