Rule-based deep reinforcement learning for optimal control of electrical batteries in an energy community

Abstract

We propose a rule-based deep reinforcement learning approach for optimal control of battery energy storage systems within a local energy community, combining interpretable rule extraction with learned control policies.

Reinforcement LearningEnergy ManagementBattery Control

BibTeX

@inproceedings{rocchetta2023rule,
  title     = {Rule-based deep reinforcement learning for optimal control of electrical batteries in an energy community},
  author    = {Rocchetta, Roberto and Nespoli, Lorenzo and Medici, Vasco and Basso, Saverio and Derboni, Marco and Salani, Matteo},
  booktitle = {Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023)},
  pages     = {639--646},
  year      = {2023}
}