A new method of blocking fault diagnosis on stator winding based on ANN

In this paper, a new method of stator winding blocking fault diagnosis based on artificial neural networks (ANN) has been proposed. At first, the characteristic of blocking fault is analyzed, and the temperature distribution in all kinds of faults is calculated, which made the one-to-one corresponding relationship mathematics model between the temperature-rise of each stator winding measuring points and the different faults founded. Then the neural network is trained, which can identify the. blocking fault position and the serious level, and the real-time example is adopted to verify the method.