RBFN based on two levels iteration cluster algorithm and its application in generator fault diagnosis

Radial basis function network (RBFN) is one of artificial neural network (ANN) applied widely. A new adaptive RBFN algorithm named as two levels iteration cluster algorithm is put forward, which can calculate automatically RBFN parameters with samples, and overcomes the conventional algorithm´s shortcoming that hidden layer neuron number must be given in advance and the different initialization method often has different cluster result and different diagnosis precision. Using practically acquired MJF-30-6 generator vibration data in three conditions of normal operation, rotor excitation winding short circuit and stator winding fault as RBFN samples, the results of verification show that the method has less learning error and higher diagnosis precision than conventional method.