Research On State Monitoring & Fault Diagnosis System For Transformer

This paper based on the JIA' XING Science & Technology Department' s key Project "The Transformer State Monitor and Fault Diagnosis through WEB Remote Visit" . Based on the detailed analysis of the large scale Transformer fault kinds and process methods, we adopt the commonly used method "Dissolved Gas Analysis in the Oil (shorten for DGA) " . Firstly, we give the detailed analysis of DGA theory, Then, according to the requirement of sample key gas in the oil, we design a data sample apparatus based on the embedded system. Finally, we develop a fault diagnosis programmer using the back BP arithmetic and general inherit arithmetic. The following paragraphs give the main contents of this paper.Chapter One firstly introduced the signification of monitoring and fault diagnosis of large scale Transformer, then discuss the transformer diagnosis trends. Based on the research background, the research task is expressed.Chapter Two analysis the DGA background and its inner theory, and task, then a system frame is show up.Chapter Three gives the signification of the sensor of monitoring and fault diagnosis of transformer system. Based on the requirement of the monitoring and fault diagnosis system, chapter analyzed the selection and the sensor A/D circuit.Chapter Four designed the data acquire equipment of the monitoring and fault diagnosis system. The data acquire equipment actualize the following function: under the operating system UCOS_II supporting, actualize the data sampling; migrate the LWIP programmer give the system having the TCP/IP protocol function, and extend the API function such as web server, the FTP custom visit.Chapter Five this chapter mainly gives a transformer diagnosis new method: GA-BP arithmetic. we want to implement the net frame and connect ratio optimize. then, emulator this diagnose programmer and using the practical data verifies it. Base on the MATLAB program environment, take several analysis methods into fault diagnosis. Using practice data for test. Result show that the diagnosis system is available.In the last Chapter, all of the works in this dissertation are summed up. Thefuture research of monitoring and fault diagnosis is prospected.