Abstract: The article discusses a method of power transformer fault diagnosis based on data fusion and extension theory, for this method a power transformer fault diagnosis system is constructed and developed by the application of wireless sensor network technology and VS.NET platform. The system applies WSN technology in wireless data acquisition of gases dissolved in power transformer oil, and can achieve real time data transmission. At the same time information fusion technology which is used to optimize the data, not only can reduce network energy consumption and ensure the accurate and objective data. For the coding rules of gas ratio method currently existed is incomplete and single, the extension method, combining quantitative and qualitative, it can make transformer fault diagnosis more refinement. Especially, the visual system can achieve the acquisition, displaying, analysis, storage and playback of gas parameters, and can be used in realtime analyzing and automatic diagnosis of transformer fault. The system of power transformer fault diagnosis can not only resolve the problem of the realtime and remote monitoring of operating personnel, but also greatly improve the performance of the transformer fault diagnosis. It provides a new direction for transformer online monitoring and fault diagnosis technology.


Key words: WSN, data fusion, extension theory, power transformer, fault diagnosis