Discrimination between Normal Current and Fault Current for Transformer Protection based on Artificial Neural Network

Transformers are vital component in power system for transfer of electricity in power system. Because of that reason protection of transformer is very much important; reliability of power system depends on transformer. Also due to high cost of the transformer and dependency of the power system on transformer, an efficient protection system has to be build which can protect transformer for any undesired condition. An ANN subroutine can build and to be used to discriminate internal fault current and magnetizing current, which can replace the traditional Fourier method and second harmonic restraining method, which are time consuming and less reliable. From study we found that now days there are certain faults, which contains second harmonic content so discrimination using second harmonic can pose a problem. This paper intends to present the advantages of ANN on other methods for discrimination between the inrush and fault condition in the transformers.