Advancing Power Grid Computing: Analysis and Synthesis of Wide-Area Data, Distributed Real-Time Hardware-In-the-Loop Simulations, and Machine Learning for Event Detection

Global Positioning Systems have allowed for precise timing of power system measurements over wide areas. This newly found capability has the potential to provide much greater insight into the operation of the power system and its response to contingencies, but few analytical techniques currently exist that provide enough robustness and trustworthiness to integrate seamlessly into power system operations. This work seeks to remedy these shortcomings in several ways. First, by thoroughly analyzing grid collected Phasor Measurement Unit (PMU) data so that data can be reliably synthesized. Second, by demonstrating a real-time Hardware-In-the-Loop (HIL) test system for analysis of high-resolution power system data complete with a comprehensive power utility visualization. Finally, a novel event detection system is developed by discovering causal network relationships in PMU data. The causal network and additional grid analytics are used in a Graph Neural Net (GNN) to detect when a power system event has occurred.

MAJOR ADVISOR: Eduardo Cotilla-Sanchez
COMMITTEE: Jinsub Kim
COMMITTEE: Rakesh Bobba
COMMITTEE: Prasad Tadepalli
GCR: Adam Schultz