Study Of Transformer Faults Diagnosis And State Evaluation Based On Information Fusion

As the essential equipment in electric power system, power transformer operation status plays an important role in safe and reliable monitoring of power system. Through detecting the transformer potential faults and abnormal status accurately and promptly, the appropriate measures can be taken, which is the assurance of troubleshooting. Recently, the transformer fault diagnosis technology is the concerned research area.Transformer fault diagnosis technology is divided into the state assessment and fault diagnosis. The transformer is divided into six parts to build the state assessment system in this article, while the amount and trends of test data is comprehensively considered, and an approach of transformer condition evaluation based on dynamic gray target is proposed in this paper. First of all, the static weight is obtained through the historical test data which is disposed by the analytic hierarchy, and then the dynamic weight is obtained through using the entropy theory to deal with the parameter changes. Secondly, a comprehensive weight is obtained, through which the static weight is amending by the dynamic weight, so that the rationality and reliability of the evaluation is improved. Finally, the model is built by constructing a dynamic gray target, and then the healthy state of transformer is obtained by the state classification strategy. The feasibility and validity is verified by actual case.When traditional evidence theory is used to analyze the fault of transformer, the reliability of the selected bodies of evidence is unscientific and subjective. In this paper, the relation between each bodies of evidence is fully considered and the reliability is ascertained through the distance measure of evidence, through which the data of Dissolved Gas Analysis and electrical tests data are combined effectively and the operation state of transformer is reflected exactly and all-sidedly. Firstly, the proposed algorithm established a similarity matrix and obtains credibility of evidence through the distance of evidence, and then preprocesses evidence through introducing credibility and construct the Basic Probability Assignment based on distance measure. Finally, preprocessed evidence is combined effectively using Dmpster rule. Through the proposed approach, the multi-characteristic signal is utilized adequately and the diagnostic accuracy is improved, and the practical engineering problems are disposed effectively with enhancing the ability to distinguish the uncertainty data. The method can quickly and accurately and more comprehensively determine the transformer running state, which is verified by comparing with traditional methods, and the transformer's security, stability and economic operational level is increased.