摘要
针对电力变压器故障的特点及传统诊断方法在变压器故障诊断中的局限性,提出了基于灰色神经网络的变压器故障诊断方法.首先将典型油中气体浓度样本集作为参考序列,挖掘出样本集中的故障信息,然后利用灰色神经网络进行变压器的故障诊断.通过大量的实例,并将诊断结果与IEC三比值法和改良三比值法的诊断结果相比较,表明基于matlab灰色神经网络的诊断方法具有更高的精确度.
Aiming at the characteristic of power transformer fault and the limitation of tra- dional fault diagnosis methods of transformer.This paper put forward fault diagnisis method of transformer based on grey neural networks.After fault diagnosis information is mined via taking collected fault samples of the power transformer as reference by the statistic analysis and fault diagnosis for transformer is accomplished according to grey neural networks..A few fault samples are analyzed in the algorithm proposed and the results are compared with those obtained by the three-ratio and amended three-ratio methods. The comparison result indicates that the algorithms proposed has high diagnosis precision.
出处
《数学的实践与认识》
北大核心
2017年第4期286-291,共6页
Mathematics in Practice and Theory
关键词
故障诊断
灰色理论
溶解气体分析
变压器
fault diagnosis
grey neural networks
dissolved gas analysis
transformer
