Research On Monitoring On Line And Fault Diagnostic System For Transformers

The on-line monitoring and fault diagnosis of transformer is that using intelligent algorithm , when the transformer is running on, to achieve on-line monitoring , fault diagnosis, the type discrimination and decision-making. It achieves the status from gas data of running transformer. When transformer is in fault, the system can diagnose it, provide the type of fault and decision-making and alert clerk in time . It will cut off the circuitry automatically in severe trouble. So it can monitor the status of transformer on-line, which greatly improves the operational stability of transformers and the service quality of electricity sectors.This paper, using embedded technology, fuzzy mathematics, neural networks, rough set theory and expert system for on-line transformer monitoring and fault diagnosis system, has important significance to improving the scientific transformer fault prediction, timeliness and the accuracy of diagnosis has important significance. The main work of this dissertation is as follows:Firstly, design the overall structure of transformer on-line monitoring and fault diagnosis system and transformer information,GPRS communicate network.Secondly, use neural networks to create a condition monitoring transformer model based on neural networks. Find the transformer fault in time in application.Thirdly, use rough sets and expert systems to establish transformer fault diagnosis and decision-making model. Which can provide reliable results and right decision-making.Fourthly, use Microsoft Visual C++ 6.0,Microsoft SQL Server2000 to design The Monitoring Terminal and Main Surface of the system. Finally, content of the paper is summed up and a tentative plan for next work is put forward.