Machine Learning (ML) techniques are gaining attention for Optimal Power Flow (OPF) problems due to the increasing complexity of energy production in modern grids.
The lack of standardized datasets and evaluation metrics has hindered progress in ML for OPF.
To address this challenge, PGLearn has been introduced as a suite of standardized datasets and evaluation tools for ML and OPF.
PGLearn provides realistic datasets, supports multiple OPF formulations, and aims to democratize access to the field by offering publicly available datasets.