Distributed Reinforcement Learning for the Management of a Smart Grid Interconnecting Independent Prosumers

In the context of an eco-responsible production and distribution electrical energy at local scale urban territory, we consider a smart grid as system interconnecting different prosumers, which all retain their decision-making autonomy defend own interests in comprehensive where rules, accepted by all, encourage virtuous behavior. this paper, present analyze model management method for grids that is shared between kinds independent actors, who respect interests, encourages each actor to behavior allows, much possible, independence from external suppliers. We here game theory model, player, investigate distributed machine-learning algorithms allow decision-making, thus, leading converge stable situations, particular Nash equilibrium. propose Linear Reward Inaction algorithm achieves equilibria most time, both single time slot across allowing maximize its