The Fault Diagnosis Model Of Transformer Which Based On The Technology Of Information Fusion

As the transformer is an important device, its operation and safety directly affects the power transmission and distribution of the entire power system. Studying the diagnosis technology of transformer is a necessary measure to protect the safety of the entire power system, reduce the accidents, promote the development of the whole national economy, and have a very important significance. By finishing of the data of the power transformer faults, making us know the relationship between the faults and symptoms clearly, and we have analyzed the feasibility of the fault diagnosis by using of the information fusion thought, Bayesian Network Learning, and Probabilistic Reasoning.In this paper we base on the technology of information fusion, and build a fault diagnosis model. The why we us the technology of information fusion is that it can handle the Multiple sources and data, thus can help us receive the more accurate and reliable diagnosis conclusion. In addition, for the existences of uncertainty factors among the fault diagnosis, especially the big electrical equipment its components and between its components existing very complex relationship, usually manifested the multiple symptoms when the different faults happened, so we adopt the Bayesian Network to handle this. Then the model we built can associate the symptoms and faults, through the learning and reasoning of the Bayesian Network, we can get the answer of every child Networks, they are the input values of the Decision Fusion Layer, it's the last floor in the diagnosis model, and its output will be the final result.