Transformer Fault Diagnosis based on RBF Neural Network

摘要

With respect to the transformer fault diagnosis, this paper proposed a fault diagnosis model based on RBF neural network, realized RBF neural network based on center vector of gauss basis function, the weight calculation method of RBF based on Kalman filtering, the training and testing algorithm of the data classification model is given. The proposed method of the transformer fault diagnosis based on RBF neural network is discussed in detail. Because the modularized structure is adopted and each sub-model is only used to recognize one fault, the difficulty of training model is reduced, it is more important that the ability and application flexibility of the fault diagnosis are improved obviously. Research results show that the proposed method has strong robustness and high accuracy.