摘要：This paper presents the application of neural networks for the non-differential protection ofsalient-pole synchronous generator against internal faults in any winding of the stator. Thedirect phase quantities and modified winding function approach has been used to simulatedifferent types of internal and external faults using electrical parameters of generators installedby utilities. The cases of all the possible types of internal faults in the stator winding have beentaken into consideration in designing the protection scheme. Multi-layer Feed-forward NeuralNetwork (MFNN) and Radial Basis Function Neural Network (RBFNN) have been trained andtested for detection, identification and classification of the internal faults based on patternclassification. The simulated fault currents in the phases as well as their parallel paths at theterminal end have been used for training and testing of both the proposed neural networks. Boththe networks are able to identify the fault signal correctly but the MFNN is more reliable, moreaccurate and faster than RBFNN in detection and classification of the fault.
