基于模糊神经网络的变压器故障诊断新方法

New Method Based on Fuzzy Neural Networks for Transformer Fault Diagnosis

基于模糊理论与神经网络理论，提出了变压器故障诊断的新方法，根据特征气体法和改良ＩＥＣ三比值法，建立了模糊神经网络诊断模型。此模型有效地处理了故障诊断中的不确定因素，并具有较强的知识获取能力，研究结果表明了这种方法的有效性和应用前景。

The Paper proposes a new method of transformer fault diagnosis on the basis of fuzzy theory and artificial neural networks (ANN) theory and sets up a fuzzy neural networks based diagnosis model according to Key-Gas method and improved IEC (International Electric Committee) three ratio method.The model can effectively handle uncertain factors in fault diagnosis and has strong ability in knowledge acquisition.Research results have shown the effectiveness and application potential of the method.