Application of genetic algorithm trained masterslave Neural Network for differential protection of power transformer

The proposed work presents the use of Artificial Neural Network (ANN) as a pattern classifier for differential protection of power transformer, which makes the discrimination among normal, magnetizing inrush, over-excitation and internal fault currents. This scheme has been realized through two separate customized Parallel-Hidden Layered ANN architectures which work in Master-slave mode. The Back Propagation Neural Network (BP) Algorithm and Genetic Algorithm (GA) are used to train the multi-layered feed forward neural network and their simulated results are compared. The neural network trained by Genetic algorithm gives more accurate results (in terms of mean square error) than that trained by Back Propagation Algorithm. Relaying signals under different fault conditions are obtained by simulating the system using MATLAB Simulink and SimPowerSystem toolbox. Simulated data are used as an input to the algorithm to verify the correctness of the algorithm. The GA trained ANN based differential protection scheme provides faster, accurate, more secured and dependable results for power transformers.