Application of a novel parallel variable weight decision tree algorithm in transformer fault diagnosis

Power transformer is the key equipment of the power system. In order to test and monitor transformers' operating conditions, a variety of fault diagnosis technology has been applied. However, these method may show some weakness such as time-consuming and over-fitting. In order to solve the problems, the paper proposes a new parallel variable weight decision tree algorithm for fault diagnosis. The experimental results based on the fault data-sets prove that the method has the advantages of higher classification precision and better fitting precision which can guarantee that the fault can be diagnosed precisely, so as to ensure the normal operation of the transformer.