Generator Rotor Winding Inter-Turn Short Circuit Intelligent Fault Diagnosis

The generators plays a vital role on power system's security and stability of operation and the rapid development of power system protection of the generator has been put forward new demands. Studying of the Generator Fault Diagnosis Technology to ensure power system security has become an important means of reliable operation. The generator protection technologies will develop towards computerization, networking and intelligent integration trend.The generator rotor winding inter-turn short circuit is a common electrical fault. For a long time, the rotor winding inter-turn fault detection for early diagnosis and on-line at home and abroad are concerned about the issue. This paper firstly research on generator rotor winding inter-turn short-circuit failure mechanism, and analyzes of the failure of the distortion of electromagnetic fields and electrical changes in the amount of features, and briefly compare the advantages and disadvantages of traditional methods. Secondly, the paper studies the neural network in fault diagnosis applications. By using the fault simulation experiments and built a sample of neural networks, it not only effectively identifies on the rotor winding inter-turn short-circuit fault, but also rightly determine the severity of failures. Then by using double-dynamic-means clustering analysis of RBF neural network, it can diagnoses rotor winding inter-turn short circuit fault. Finally, the particle swarm optimization of neural network is applied to the rotor winding inter-turn short-circuit fault diagnosis.A lot of examples shows the two algorithms proposed in the paper not only can identify the rotor winding short circuit fault, but also has their precision and accuracy compared with traditional algorithm significantly improved.