A physics-guided RL agent that uses line outage distribution factors to choose practice actions during training achieves longer average blackout survival on Grid2Op benchmarks than random exploration.
Hill, and Gustaf Olsson
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RL for Mitigating Cascading Failures: Targeted Exploration via Sensitivity Factors
A physics-guided RL agent that uses line outage distribution factors to choose practice actions during training achieves longer average blackout survival on Grid2Op benchmarks than random exploration.