ANN-Based Approach for Transmission Line Fault Detection and Localization

The application of artificial neural networks (ANNs) for the identification and classification of electrical power transmission line faults using an IEEE 9 bus system is explored in this research.. The proposed method uses three-phase current and voltage measurements taken from one end of the transmission line as input parameters. A feed forward neural network, trained through the back propagation algorithm, is used in fault detection. To enhance the understanding and effectiveness of the model, two ANN tools are used. During training, ANN learns the input and output parameter mapping and provides results as a function of training performance and also plots for further analysis. Fault classification is done to study the behavior of each of the three phases separately. Besides, the study was carried out with a complete analysis by varying the number of hidden layers to identify the optimal configuration of the neural network. Simulation outputs indicate that the proposed ANN-based method identifies and classifies faults in transmission lines effectively and accurately. The main focus lies in method for effective, reliable and secure result of the system using artificial neural network. The explore of ANN method with data extracted from main bus system. The voltages of three phases and their corresponding currents are being taken as input for ANN. The training, testing and validation of dateset are done to train data as per target data and the accuracy obtained as 98.7%. The simulated results are shown effective result of ANN performance.