Application of Modified Combinatorial Radial Basis Function Neural Network in Fault Diagnosis of Power Transformer

This paper proposes a RBF Neural Network and the conception of combinatorial diagnosis and applies them in fault diagnosis of power transformer. The RBF Neural Network is prior to BP Neural Network in the ability of approach , the ability of classification and the rate of train. After the input data is processed and the training method is modified, combinatorial radial basis function neural network has a rapid train rate, high convergence precision and high diagnosis veracity in fault diagnosis of power transformer.