Proper

Abstract

Transformer is one of the key equipments in power system, and its running state is closely related to the safe and reliable operation of power system. With the development of ultra-high voltage power transmission technology and the expansion of power system, transformer fault caused great harm to the power system will be. In this thesis, the gas dissolved in power transformer oil and fault type relationship on the basis of diagnosing feature around the fault diagnosis technology of power transformer selection, fault diagnosis and fault prediction is studied in three aspects. A hierarchical fault diagnosis model is proposed based on the characteristics of the fault diagnosis and the effective information amount of the fault diagnosis. According to the size of effective information, overheating and discharge layer C2H2 and CO2, and in the low temperature high temperature overheating overheating layer selected C2H2, CO, CO2, optimal diagnostic characteristics of high energy discharge and low energy discharge layer is C2H2, CO, CO2. Building gas photoacoustic spectroscopy detection system, using spectral measurement technology to detect the concentration of gas in photoacoustic cell. According to the proposed optimization model, the error coefficient of transformer winding hot spot temperature curve is significantly smaller than that of other methods.