Improved Fault Classification Method in Transmission Line based on K-means Clustering

An intelligent fault detection and classification for abnormal current data of transmission lines is proposed. According to the monthly report and the abnormal condition information stored in the system, four kinds of abnormal state data are chosen. By extracting three fault features of data, combining k-means algorithm and TSNE dimensionality reduction algorithm, fault categories can be clearly and intuitively distinguished, and the purpose of intelligent alarm for abnormal state current is realized.