MO-ADRL: Multi-Objective Adaptive Deep Reinforcement Learning for Intelligent Grids with High Renewable Penetration
Source code for the MO-ADRL framework: intelligent-grid simulator, Temporal Transformer forecaster, actor-critic agent with adaptive multi-objective reward weighting, prioritized experience replay and uncertainty-aware exploration, together with rule-based, PSO and GA baselines and convergence-analysis scripts.
Puneet Sharma, B.N. Srinivasarao, Dr. K.Hemachandra Reddy et al.
· Zenodo (CERN European Organi... · 0 citations