Numerical Differential Protection of Power Transformer using GA Trained ANN

This paper presents the use of ANN as a pattern classifier for differential protection of power transformer, which makes the discrimination among normal, magnetizing inrush, over-excitation, external fault and internal fault currents. The Back Propagation Neural Network Algorithm and Genetic Algorithm are used to train the multi-layered feed forward neural network and simulated results are compared. Comparing the simulated results of the above two cases, training the neural network by Genetic algorithm gives more accurate results (in terms of sum square error) and also faster results than Back Propagation Algorithm. The GA trained ANN based differential protection scheme provides faster, accurate, more secured and dependable results for power transformers.