A PMU-Based Data-Driven Approach for Classifying Power System Events Considering Cyberattacks

Abstract

Transient event identification is essential for power system operation and situational awareness. The increased penetration of the high sampling frequency phasor measurement units (PMUs) enables using PMU data to analyze power system events, and thus enhances power system visualization, monitoring and control. At the same time, the risks associated with cyberattacks on power systems increase. A malicious cyberattack on PMUs, aiming at generating fake transient data, may lead to incorrect actions that jeopardize system reliability. Therefore, it is critical to distinguish between fake data and real data when analyzing transient events. Utilizing PMU measurements, this article develops a data-driven approach, based on text-mining methodologies, for classifying transient events and identifying fake events caused by false data attacks. The developed methodology provides credible information regarding the cause of various events, and facilitates postevent decision-making to prevent potential cascading failures. Case studies, performed on the IEEE 30-bus and IEEE 118-bus systems, show that the developed approach is efficient in classifying false data and identifying different transient events regardless of the system topology, loading conditions, or the placement of PMUs.