The operation of modern power grids has become increasingly challenging with the transition to renewable energy sources. As a consequence, the transmission system operator of France has launched a series of competitions, called L2RPN [1], to explore novel methods to operate power grids. During my work at Hitachi Energy, I explored a novel control strategy based on optimal power flow, reinforcement learning, and imitation learning for my master’s thesis. In our simulation, the method managed to prevent 86% of all blackouts and resulted in 72% longer grid operation time and was presented at the 5th Hitachi AI Conference 2024.
The final controller performs a combination of redispatching (power plant adjustments), curtailment (limiting renewables), battery storage power injections, and load shedding (partial shutoff from the grid) in the worst case.
[1] Marot, Antoine, et al. “Learning to run a power network challenge: a retrospective analysis.” NeurIPS 2020 competition and demonstration track. PMLR, 2021.