Based On Neural Network Analysis Of The Oil-immersed Transformer Fault Diagnosis

Grid in the power system to the responsibility of the transmission, substation, distribution of three parts, and the main transformer is the heart of the power grid, It is used frequently in the power grid, but it is also an accident more equipment, Its operation of the relationship between the power grid safe operation and economic benefits. When the main transformer failure occurs, will impact and threat to the safe and stable operation of power grid, So the use of the main transformer fault diagnosis to reduce or avoid the occurrence of accidents is very important.According to the study of the main transformer failure phenomena, causes and failure detection technology, Further analysis of one of the most important, the most commonly used methods-- In the oil gas chromatography. By studying the composition of the gas,gas properties,gas rate, The main transformer internal fault divided into two categories, overheating and discharge breakdown,With characteristic gas method and three ratio method to judge the fault. By means of Shangqiu fault several main transformer actual case analysis, describes how to oil by gas chromatography in the troubleshooting and diagnostic analysis; Other detection technology are introduced at the same time found the failure instance and failure analysis method. Finally through to the neural network learning, Using the MATLAB software is established based on BP neural network model of transformer fault diagnosis. And through two kinds of improved BP algorithm to fault gases as input, the actual production data of gas generation, when a few cases, through the simulation on the failure data for further analysis, Get the ideal output and a feasibility analysis of the method.