As generating unit, synchronous machines play very important role in safety and stability of power systems. As shorted rotor winding is a usual fault in the faulty of synchronous machines, the paper studies the diagnose method, which combines the traditional method with the intelligent method.Firstly, the theory and character of the shorted rotor windings is analyzed. A new intelligent method, i.e. Genetic Programming(GP), is introduced. It is better than other method in symbol regress, so the method is used to establish the model of magnetization curve. The advatages of the method is shown in case studies.Secondly, Artificial Neural Network(ANN) is introduced. BP(Back Propagation) network is used to simulate the magnetization curve of generator. The results with GP and ANN are compared.At last, the importance of fault diagnosability is analyzed. Then the fault diagnosability of two intelligent methods is analyzed qualitatively. |