Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Grid Modelling topic
No spam. Unsubscribe anytime.
Brookhaven researcher says AI ‘foundation models’ can evaluate billions of grid scenarios in minutes
Summary
Hendrik Hammond of Brookhaven National Laboratory described grid foundation models trained on electrical variables that can rapidly evaluate billions of scenarios, producing speedups reported as high as 1,000× and shortening studies that once took months to minutes.
Hendrik Hammond, chief AI scientist at Brookhaven National Laboratory and a professor at Stony Brook, told the plenary that the electric grid’s size and complexity require new modeling approaches. He said conventional scenario analysis examines only a few possibilities while AI foundation models trained on voltages, power and frequency enable many more scenarios to be evaluated quickly.
Hammond said the group has already seen speedups "up to 1000," and that simulations that formerly took months to complete can now be run in minutes, which he argued will improve planning and security for the national grid. "With AI, we can remove the blindfolds," Hammond said, describing the approach as a community effort involving utilities, hyperscalers and universities.
