Development of Artificial Neural Network Modules for Transmission Line Fault Protection

Abstract

Transmission line faults can be classified into various categories based on nature of bus voltage and line fault currents. Monitoring of these two factor can become very useful for design of protective systems in power system. The impact of fault resistance and fault inception angle have to be considered, while analyzing type of faults occurring on transmission line. For the said purpose an approach of Artificial Neural Network(ANN) is adopted. The ANN developed using software tools has been trained and used for detecting and predicting the type of fault in a large interconnected transmission system. In the software tool the symmetrical and unsymmetrical faults are simulated and analyzed. The data produced from this software are used for the training and testing purpose. Faithful fault detection under noisy data scenario and /or missing measurement data is challenging task and it is being taken up as the task. The results of the ANN outcome are embodied in the thesis.