Researchers in the Electrification and Energy Infrastructures Division are developing automated methods for detecting abnormal equipment operation and activity in the electric grid, thwarting cyber intrusion, and securing grid communications. These efforts enhance the safety and reliability of the nation’s critical infrastructure, including its power supply.
Researchers develop and validate alternative timing architectures to augment the use of GPS time for synchronization in the electric grid, as well as offline channels for grid communications. These innovative methods move vital functions farther from more accessible platforms that could be susceptible to tampering. ORNL researchers have invented new methods for validating automated communications among grid equipment using blockchain frameworks, quantum solutions, and analysis algorithms enriched by machine learning.
EEID researchers are also creating and managing platforms for datasets that enable utilities to train machine learning, enabling autonomous power system operations. ORNL maintains a publicly available Grid Event Signature Library as a repository for grid data, including waveforms of abnormal incidents such as electrical arcing. The data can be used in algorithm development, allowing rapid recognition of operating problems that could result in blackouts or fires. Other EEID researchers take the next step, creating the signal processing and algorithms for AI-assisted analysis of real-world grid sensor data in real time. This enhances situational awareness, increasing the resilience of the electric grid while reducing maintenance costs and downtime.
