Power Transformer Differential Protection based on Artificial Neural Network and Wavelet Transform
Differential Protection, Artificial Neural Network, Wavelet Transform
The proposed method uses the wavelet transform to detect the beginning of some disturbance affecting the transformer, either internal or external. Whereas ANN will correctly classify the type of disturbance (internal fault, external fault and transformers energization), quickly and effectively. Therefore, RNA make the function the classifier of patterns that has the same function of differential protection of transformers, with a high efficiency in the detection of internal faults, thus avoiding inadequate relay operations in situations of internal fault and transformer energization. In addition, the proposed method has an additional function, which is the classification of faults internal of transformer, which can assist in the repair and maintenance of the transformer. Therefore the proposed differential protection scheme can work in conjunction with the other system protections by sending warning signals, such as for external fault conditions of transformer, thus making the protection more reliable and intelligent.