Fault Identification and Classification in Transmission Line by ANN Technique Using Levenberg-Marquardt Algorithms

Abstract

It is very important to identify and classify the nature of transmission line fault for reliable power flow
and efficient operation of power system. It is also important for relaying decision and auto reclosing requirements. The
introduction of pattern recognizer has provided great progress in the power system protection. Artificial Intelligence (AI)
such as artificial neural network (ANN) can be utilized for recognition and classification of fault in transmission line.
The ANN method requires three phase voltage and current at fault point. These signals are applied at input of the ANN
pattern recognizer. Training, validation and testing are the three essential steps employed in ANN. The back propagation
learning technique is used along with Levenberg-Marquardt algorithms.