首页> 外文会议>2013 IEEE 1st International Conference on Condition Assessment Techniques in Electrical Systems >Virtual instrument based fault classification in power transformers using artificial neural networks
【24h】

Virtual instrument based fault classification in power transformers using artificial neural networks

机译:使用虚拟神经网络的基于虚拟仪器的电力变压器故障分类

获取原文
获取原文并翻译 | 示例

摘要

Inrush currents in power transformers are detected based on magnitude of second harmonic component. To avoid the harmful effects of inrush, amorphous core is widely used in recent days. Transformers with amorphous core cause low magnitude inrush current and hence the second harmonic of inrush current is comparable with that during internal faults. This increases the chances for relay mal operation when classical techniques of discriminating inrush from other faults are used. To overcome this, advanced signal processing techniques like wavelets, S-transform, H-transform and pattern recognition tools like fuzzy logic, neural network, support vector machine etc. are being used in recent days. A combination of wavelets and neural network is found to give satisfactory solution to the above problem. In this paper, a comparative study using different mother wavelets along with different activation function is made to enhance the performance. Virtual instrument is used to demonstrate the method of fault classification.
机译:基于二次谐波分量的大小来检测电力变压器中的浪涌电流。为了避免涌入的有害影响,近来非晶核被广泛使用。具有非晶铁芯的变压器会产生低幅度的浪涌电流,因此浪涌电流的二次谐波与内部故障期间的谐波相当。当使用区分浪涌和其他故障的经典技术时,这增加了继电保护操作的机会。为了克服这个问题,近来正在使用诸如小波,S变换,H变换之类的高级信号处理技术以及诸如模糊逻辑,神经网络,支持向量机等的模式识别工具。发现小波和神经网络的组合可以为上述问题提供令人满意的解决方案。在本文中,使用不同的母小波以及不同的激活函数进行了比较研究,以提高性能。用虚拟仪器演示了故障分类的方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号