Partial Discharge Pattern Recognition Algorithm Based on Sparse Self-coding and Extreme Learning Machine

HE Jin; CAO Meng; WANG Wei; ZHU Xuliang; XING Xiangshang; CHEN Rong

doi:10.13296/j.1001-1609.hva.2018.11.043

您当前的位置：

首页 >

文章列表页 >

Partial Discharge Pattern Recognition Algorithm Based on Sparse Self-coding and Extreme Learning Machine

更新时间：2025-04-26

- Partial Discharge Pattern Recognition Algorithm Based on Sparse Self-coding and Extreme Learning Machine
- High VoltageApparatus Vol. 54, Issue 11, Pages: 295-300+306(2018)
- 作者机构：
  
  国网天津市电力公司电力科学研究院
- 作者简介：
- 基金信息：
- DOI：10.13296/j.1001-1609.hva.2018.11.043
  CLC： TM855
- Published：2018
- 稿件说明：
移动端阅览
[1]何金,曹梦,王伟,朱旭亮,邢向上,陈荣.基于稀疏自编码和极限学习机的局部放电模式识别[J].高压电器,2018,54(11):295-300+306.

HE Jin, CAO Meng, WANG Wei, et al. Partial Discharge Pattern Recognition Algorithm Based on Sparse Self-coding and Extreme Learning Machine[J]. High voltageapparatus, 2018, 54(11): 295-300+306.
[1]何金,曹梦,王伟,朱旭亮,邢向上,陈荣.基于稀疏自编码和极限学习机的局部放电模式识别[J].高压电器,2018,54(11):295-300+306. DOI： 10.13296/j.1001-1609.hva.2018.11.043.

HE Jin, CAO Meng, WANG Wei, et al. Partial Discharge Pattern Recognition Algorithm Based on Sparse Self-coding and Extreme Learning Machine[J]. High voltageapparatus, 2018, 54(11): 295-300+306. DOI： 10.13296/j.1001-1609.hva.2018.11.043.

摘要

传统的局部放电模式识别方法识别正确率低

或者训练时间长。文中提出了一种新的局部放电模式识别算法

绘制局部放电信号的PRPS图谱作为输入数据

采用基于稀疏自编码器（sparse autoencoder

SAE）实现对PRPS图谱的特征提取和降维

得到能高度表达原始数据的低维特征空间。利用极限学习机（extremelearning machine

ELM）网络作为分类器

实现对局部放电的分类。利用实验得到的数据样本测试该算法

结果表明该算法不仅模式识别正确率高

并且训练速度快。

Abstract

To improve the recognition accuracy and shorten training time of traditional partial discharge（PD） pattern recognition method

a new PD pattern recognition method is proposed. The PRPS map of the PD signal is extracted asthe input data. The feature extraction and dimension reduction of the PRPS map are realized based on the sparse auto-encoder（SAE）to obtain low-dimensional feature space that can express the original data sufficiently. The extremelearning machine（ELM）network is used as a classifier to realize the classification of PD. Experimental results show that the proposed algorithm has a high accuracy of pattern recognition with fast training speed.

关键词

Keywords

references

Views

189

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Statistical Fingerprint Analysis for Capacitor PD under DC

Research on a Cascade Partial Discharge Locating Method for High Voltage Switchgear Cabinet

Distance-based Methods for Pattern Recogition of Discharge

A New Method of Reconstructing PD Signal from Self-correlation Function

Interference Immunity of UHF Method for Partial Discharge Detection in GIS

Related Author

LIU Hua-chang

WU Chun-yan

ZHANG Hangwei

CUI Jiangjing

LIANG Yuxiong

YAN Yuan

LI Hongjie

WU Hongxiao

Related Institution

The Institute of Electronic Engineering of CAEP

Zhaohui School of CAEP

School of Electrical Engineering, Xi'an Jiaotong University

Zhuhai Power Supply Bureau, Guangdong Power Grid Corporation

清华大学北京 100084

⁰