Faulted Line Identification in Power Network using Unsupervised Machine Learning

Identification of faults in a power grid is crucial for maintaining stability of the network. Fault identification can help in operation of relays in the network and also provide real time information to network operators. In this paper, a machine learning based algorithm is proposed for faulted line identification in the network. Wide area measurement system (WAMS) provide synchronized data from different location in the network. This data is used for the proposed method. Principle component analysis (PCA) and k-means algorithm are used for faulted line identification in a power grid. The proposed method can identify fault conditions in the network and discriminate them from other stressed situations such as power swing. The method has been tested for the IEEE-9 bus network. The performance of the method is also tested with a renewable energy source in the network. The results obtained from PSCAD simulations highlight the accuracy of the proposed method in faulted line identification.