Fault Recognition of On-Load Tap-Changer Based on Improved BP Neural Network

被引：1

作者：

Jiang, Chunbing ^{[1
]}

机构：

[1] North China Elect Power Univ, Sch Elect & Elect Engn, Beijing, Peoples R China

来源：

2017 3RD INTERNATIONAL CONFERENCE ON COMPUTATIONAL SYSTEMS AND COMMUNICATIONS (ICCSC 2017) | 2017年

关键词：

Fault recognition; On-load tap-changer; Neural network; Variable learning rate;

D O I：

10.23977/iccsc.2017.1005

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

On-load tap-changer (OLTC) is an important part in power transformer, which is used to change the transformation ratio by switching contact form one winding tap to another without interrupting the load. It is necessary to make sure the effective running of OLTC. Thus, a fault recognition method for OLTC based on improved BP neural network is proposed in this paper, according to the working characteristicsof OLTC. The feature of this improved neural network is that the learning rate is variable. Experimental results on 2000 test samples from offline AC test device have showed the effectiveness of the proposed fault recognition method.

引用

页码：24 / 28

页数：5

共 11 条

[1]

[陈小玉 Chen Xiaoyu], 2012, [计算机仿真, Computer Simulation], V29, P318

[2]

Dong Qiguo, 2001, FAULT DIAGNOSIS ELEC, P10

[3]

Huang Deshuang, 1996, NEURAL NETWORK PATTE, P105

[4]

Li JP, 2006, PROCEEDINGS OF 2006 INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND CYBERNETICS, VOLS 1-7, P809

[5]

Lu Gan-yun, 2005, Proceedings of the CSU-EPSA, V17, P19

[6]

Ming Meng, 2003, P CSEE, V23, P82

[7]

Wang Xiao-xia, 2008, East China Electric Power, V36, P112

[8]

Xu Wen, 1997, Proceedings of the CSEE, V17, P109

[9]

Yao Zhisong, 2002, PRACTICAL MANUAL LOA

[10] Support vector machines-based fault diagnosis for turbo-pump rotor [J].

Yuan, SF ;

Chu, FL .

MECHANICAL SYSTEMS AND SIGNAL PROCESSING, 2006, 20 (04) :939-952

← 1 2 →