Reinforcement Learning using the Actor-Critic framework for the L2RPN challenge (https://l2rpn.chalearn.org/ & https://competitions.codalab.org/competitions/22845#learn_the_details-overview). The agent trained using this code was one of the winners of the challenge. The code runs on the pypownet environment (https://github.com/MarvinLer/pypownet)…

The implementation of NeurIPS_2020_L2RPN_Track1(Robustness) and Track2 (Adaptability) Competition

Research implementation of Meta-Learning (MAML) in Multi-Agent Reinforcement Learning (MARL) for power grid control using L2RPN. This project explores how meta-learning can improve agent adaptation across varying grid topologies and operational scenarios, aiming for safer and more efficient grid management.

Source code for the papers: RL for Mitigating Cascading Failures: Targeted Exploration via Sensitivity Factors (NeurIPS) / Blackout Mitigation via Physics-guided RL (IEEE TPS). Built with TensorFlow.

Research on Hierarchical Multi-Agent Reinforcement Learning for Power Grid Topology Control (L2RPN). This project explores hierarchical MARL architectures, regional agents for stabilizing power grids using the Grid2Op environment.