Power Transformer Faults Diagnosis Approach Based On Bayesian Network

Power transformers belong to the most important components of power generation and transmission system. In this paper, the main fault classes, fault monitoring techniques and fault symptom variables are analyzed synthetically. The new transformer fault diagnosis approach based on Bayesian network is proposed, which has been proved to be efficient in dealing with problem of uncertainty. Two forms of Bayesian network diagnosis model are established and they are testified to be precise and reliable. Several incomplete-data learning methods are applied to solve the transformer data absent, whose performance and shortcoming are explained through the experiment. Finally, a transformer fault diagnosis system is established based on the Bayesian network.