Electrical power plants require fault detection and operation of protection equipment in minimum possible time to remain stable. The faults on electrical power system transmission lines should be first detected and classified correctly, and repaired in the least possible time. The protection system used for a transmission line can also initiate the other relays to protect the power system from outages. A sound fault detection system provides a fast and secure relaying operation. Once trained on reliable data, an ANN-based fault detection system can help provide fast predictions for the presence of a fault and classify what kind of fault it is. Data for training the ANN was obtained from Kaggle.
Fault Detection ANN Model:
Fault Classification ANN Model:
For Fault Detection, the model predicted the faults with an accuracy of 99.41% and for Fault Classification, the model was able to classify faults with an accuracy of 99.42%.