Generator Event Detection from Synchrophasor Data Using a Two-Step Time-Series Machine Learning Algorithm

Abstract

The electrical faults of generators on an electrical grid, i.e., generator events (GE), must be detected efficiently when they occur, as these events can propagate to the rest of the grid in a cascading manner, leading to outages and wide-area blackouts. Many reasons exist that give rise to these faults, but at its most fundamental, they constitute an inability of a generator to match the grid usage requirements. In this paper, we present an efficient algorithm to accurately identify the occurrence of generator events within an electrical grid, using the monitoring data obtained from phasor measurement units (PMUs). Specifically, we have developed a machine learning algorithm that takes as input PMU data, and subsequently flags instances where a GE had taken place. The detection is performed in near real-time with the help of a standard off-the-shelf processing unit. Furthermore, we set out to create electrical fault maps that demarcate the progression of the event as it takes place. Experimental results show that our algorithm can accurately and efficiently identify the occurrence of a GE. In addition, we are also able to report a fault network map, which provides a powerful tool for troubleshooting.