Fault Detection Classification and Location of Transmission Line Using AI Techniques

This work presents an intelligent protection scheme in high voltage transmission line using artificial neural network (ANN). The objective is to analyse the performance of artificial neural network based relay to detect, classify and locate faults in transmission line using feed-forward network along with backpropagation algorithm. The input features of the neural network are fundamental frequency voltage and current magnitudes extracted by discrete- Fourier transform. One full cycle DFT has been used extract RMS of fundamental frequency components of voltage and current signals. These samples are then normalized between range of 0 to 1 as a ratio of post fault to pre- fault values of voltage and current signal so as to suit input of ANN. The inputs will be applied to ANN pattern recognizer. The artificial neural network based relay is very effective to locate faults on transmission lines and achieve satisfactory performances