Multilevel Information Fusion DGA Diagnosis Method for Power Transformer

Different types of failure occur in the running transformer, always goes with different warning signs. Fault diagnosis is to analyze the information, and then judge the nature of the fault severity, and found the transformer faults and abnormal state parts, At present, with the application of new detection methods such as on-line monitoring and live detection of transformers, more and more data representing real-time state information of transformers and more and more information sources are available. However, on the one hand, the characteristics of the diagnostic methods are characterized by a single type of characteristic information, such as the ratio of gas content or content, which leads to weak anti-interference ability and easy to cause errors. On the other hand, the characteristics of diagnostic methods may have different characteristics, and the lack of integration and comprehensive analysis among information affects the analysis and utilization of characteristic information. In general, the current use of information remains at a lower level. To this end, this paper introduces the information fusion technology, and puts forward the diagnosis new criterion to be effectively integrated and rationally stratified in order to improve the reliability of the diagnosis.