Q-Learning Based Energy Management for Microgrids: Internalizing SOC Constraints for Fast Online Inference
While traditional day-ahead economic scheduling provides a foundational baseline for microgrid energy management, the inevitable transition toward real-time, dynamic control necessitates algorithms with extreme computational efficiency. This paper presents a microgrid optimization strategy predicated on the Q-learning...