三种变压器故障诊断方法比较研究

变压器常规诊断的油中溶解气体的三比值法(IEC),在目前变压器故障诊断中有着广泛的应用,许多人工智能技术在电力变压器故障诊断中的应用都是基于IEC法进行研究的,并在此基础上提出了基于神经网络和模糊神经网络的变压器故障诊断专家系统。该文将传统IEC三比值法、神经网络法和基于IEC三比值法的模糊神经网络法进行了比较。本文结合实例对这三种诊断方法进行了分析与探讨,指出了传统的IEC三比值法的不足,神经网络法和基于IEC三比值法的模糊神经网络法的优点等。

The Comparison Research on Three Kinds of Fault Diagnosis of Power Transformer

The fault of electric power transformer and the IEC (three ratio method) used in the routine diagnosis electric power transformer are introduced. The IEC technology is used frequently in electric power transformer fault diagnosis, and now many artificial intelligence technology were used in electric power transformer fault diagnosis are based on the IEC technology. On this condition the paper put forward the applicability of the transformer fault diagnosis expert system based on ANN and FNN. An artificial neural network theory, the artificial neural network based on fuzzy logic theory, and also conventional IEC three ratio method are applied to fault diagnosis of transformer. There will be comparison among the three method. And data pretreatment of the neural network is achieved by fuzzy math theory, which enhance the veracity and reliability of diagnosis with conventional IEC three ratio method in power transformer.