Research on Transformer Fault Diagnosis based on Neural Network
Shengnan Tang · Microcomputer Information · 2010
This paper adopts a method of chromatogram analysis on composition and concentration of transformer oil gas to judge the potential faults in fault diagnosis of power transformer.As the known fault types are limited,this paper puts forward an improved RBF neural network algorithm,which is used to solve the problem of automatic knowledge-base of fault types updating.Experiment finds that the application of this algorithm will increase the efficiency of diagnosis and strengthen the automatic updating and maintenance of knowledge-base of fault types.