This paper presents an alternative approach using the differential logic associated to Artificial Neural Networks for protection of power transformers. Neural networks are used for distinguishing internal faults from inrush currents produced during the transformer energization. The EMTP-ATP software has been chosen as the computational tool to simulate the power transformer in order to generate data to train and test the ANNs. A complete differential protection software was implemented in FORTRAN and encouraging results related to the application of the new method are presented.
System Protection; Power Transformers; Differential Protection; Artificial Neural Networks (ANNs)
