Fault Identification & Classification in an Interconnected Power System Network using Machine Learning

Power Systems consist of three generalized phases namely generation, transmission, and distribution. It is the backbone of any electrical system. In any such network, faults create the major threat of damage and collapse of the entire grid if left unchecked. Thus, knowing the nature of the fault(s) occurring and determining the location of the same is a major challenge to some of our outdated grids. As a result of such a situation, even the right measures to be taken are delayed and millions are left in the dark with no electricity. The objective of this project is to predict and segregate the faults, occurring on the lines in various positions in an interconnected power system network. The preferred is a single-phase IEEE 5 Bus System. The idea is to model the same using MATLAB/ SIMULINK for fault feature extraction. The obtained data is used to create a master dataset. Machine learning models are trained and tested to obtain the classification results. The results are verified using the confusion matrices. In the end the accuracy scores of the algorithms are compared.