Grid2Op

AI policy distillation creates inspectable grid control models

Researchers have developed a method to distill complex deep reinforcement learning policies for power grid operation into more compact and interpretable tree-based models. These distilled models, a decision tree and a r…