A Machine Learning Approach for Line Outage Identification in Power Systems

Abstract

This paper addresses power line topology change detection by using only measurement data. As Phasor Measurement Units (PMUs) become widely deployed, power system monitoring and real-time analysis can take advantage of the large amount of data provided by PMUs and leverage the advances in big data analytics. In this paper, we develop practical analytics that are not tightly coupled with the power flow analysis and state estimation, as these tasks require detailed and accurate information about the power system. We focus on power line outage identification, and use a machine learning framework to locate the outage(s). The same framework is used for both single line outage identification and multiple line outage identification. We first compute the features that are essential to capture the dynamic characteristics of the power system when the topology change happens, transform the time-domain data to frequency-domain, and then train the algorithms for the prediction of line outage based on frequency domain features. The proposed method uses only voltage phasor angles obtained by continuous monitoring of buses. The proposed method is tested by simulated PMU data from PSAT [1], and the prediction accuracy is comparable to the previous work that involves solving power flow equations or state estimation equations.