基于BP神经网络和改良三比值法的变压器故障诊断方法
A Fault Diagnosis Strategy for Oil-immersed Transformer Based on BP Neural Network and Improved Three-ratio Method
田军 1朱占春 1孙建鸿 1魏云龙 1夏星 1钟建伟 2梁会军2
作者信息
- 1. 国网湖北省电力有限公司宣恩县供电公司,湖北 宣恩 445500
- 2. 湖北民族大学智能科学与工程学院,湖北 恩施 445000
- 折叠
摘要
变压器在电网中起着能量传输作用,其稳定性直接决定了电网供电质量.从改良三比值法出发,提出了基于BP神经网络的油浸式变压器在线故障诊断方法.该方法以变压器油中溶解的典型特征气体种类和浓度出发,给出三比值法及改良三比值法的原理,通过改良三比值法形成BP神经网络的训练数据,并以此对典型油浸式变压器故障类型进行分类.实验结果表明,方法对典型油浸式变压器故障可以进行有效分类,为变压器故障诊断提供了有效支撑.
Abstract
In this paper,inspired by the improved three ratio method,an online fault diagnosis method for oil immersed transformers based on BP neural network is proposed.The training data for BP neural network is obtained by analyzing the typical characteristic gases dissolved in transformer oil,and typical fault types of oil immersed transformers are classified.The experimental results indicate that the proposed method can effectively classify typical faults in oil immersed transform-ers,which provides an effective support for transformer fault diagnosis.
关键词
故障诊断/油浸式变压器/改良三比值法/BP神经网络Key words
fault diagnosis/oil-immersed transformer/improved three-ratio method/BP neural network引用本文复制引用
出版年
2024
NETL
NSTL
万方数据