“Researchers can search different variables, like the type of sensor or a description of the event – for example, a waveform that represents a blown transformer and its effect on transmission line voltage,” said ‘Grid Signature Event Library’ manager Aaron Wilson (pictured).
It has been created in response to the increasing use of data-driven analysis and design using neural networks, for example, as a complement physics-based techniques.
“Machine learning can be trained to recognise waveforms that provide an early warning of equipment malfunction, enabling power system operators to prevent black-outs, wildfires and damage to the power grid,” according to ORNL.
The anonymised data, which includes voltage, current and frequency information has been been collected from operating equipment by utilities and research institutions.
A description of the library and its use has been written by ORNL with the Pacific Northwest and Lawrence Livermore National Laboratories, and published by the IEEE as ‘The Grid Event Signature Library: An open-access repository of power system measurement signatures‘, or go straight to the Grid Event Signature Library here.
