Method for Insulation Defect Pattern Recognition of Gas Insulated Switchgear based on Support Vector Machine Algorithm

DOI

10.2991/icaees-15.2015.103How to use a DOI?

Keywords

partial discharge (PD); gas insulated switchgear (GIS); support vector machine (SVM); pattern recognition.

Abstract

As partial discharge (PD) can reflect the type of insulation defects in a gas insulated switchgear (GIS) and the damage of GIS caused by different types of discharge varies evidently, identifying the type of discharge correctly is of significant value in ensuring the safe and reliable operation, assessing the insulation condition and making a rational maintenance strategy for GIS. In order to study the characteristics of PDs triggered by different defects in GIS, we designed four kinds of typical discharge defects to simulate the insulation defects that may occur in a GIS. To describe the typical characteristics of PDs, eight statistical characteristic parameters were extracted from the ultra-high-frequency signals acquired by experiments. A classifier that can achieve quaternary classification was constructed based on the support vector machine (SVM) algorithm, and then the PD type was identified by voting method. Experimental results show that the proposed method possesses a high recognition accuracy and can effectively identify four typical PDs in GIS.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Jun Xiong
AU  - Sen Yang
AU  - Guangmao Li
AU  - Xiaogui Wu
PY  - 2015/07
DA  - 2015/07
TI  - Method for Insulation Defect Pattern Recognition of Gas Insulated Switchgear based on Support Vector Machine Algorithm
BT  - Proceedings of the 3rd International Conference on Advances in Energy and Environmental Science 2015
PB  - Atlantis Press
SP  - 560
EP  - 566
SN  - 2352-5401
UR  - https://doi.org/10.2991/icaees-15.2015.103
DO  - 10.2991/icaees-15.2015.103
ID  - Xiong2015/07
ER  -