A Polynomial Features Extraction with SMOTE Model for Transmission Line Faults Detection & Classification with Gradient Boosting Algorithms

The world applications are transforming with the latest technologies in various fields. Artificial intelligence and machine learning models replaces the traditional methodologies due to their fast computing ability and high accuracy. One of the research areas in the power systems is ‘Transmission Line Fault Detection’ because of the faults in transmission line creating a great damage (financial loss) to the equipment in an electrical network and reduces the reliability of power supply. To overcome these drawbacks a fault prediction and classification system is required. This work adopt the machine learning model of Gradient boosting algorithms with SMOTE (syntactic minority oversampling technique with min-max optimization scalar ) to predict the faults in transmission line and classifying the faults into respective categories by developing an appropriate machine learning model with proper training and testing. These models predict the faults and give the information to protecting systems (circuit breakers) to protect the transmission lines and disconnect the unhealthy sections (fault area) in power system network.