模糊神经网络在变压器油中溶解气体故障诊断中的应用

基于模糊隶属度和BP神经元网络,提出了将模糊神经网络应用于变压器油中溶解气体故障诊断的方法。该方法采用了由输入层、输出层、隐含层和模糊化层组成的一种四层前向模糊神经网络,并利用模糊理论预处理数据,建立了基于模糊神经网络的变压器故障诊断模型。结果表明,该方法对变压器进行故障检剥诊断是有效的。

Application of Fuzzy Neural Network to Transformer Fault Diagnosis by Dissolved Gas-in-oil

In this paper, a fault diagnosis method on the basis of the fuzzy neural networks techniques for transformer fault diagnosis by dissolved gas-in-oil. This paper applies a four layer neural network to building the model of fault diagnosis of transformers,and also uses fuzzy mathematics to dea1 with the data in advance.The experiments prove that this method is effective and can also he applied to the fault diagnosis of other electric equipment.