This repository contains combinatorial multi-armed bandits based algorithms to optimize energy flows in smart grids. There are following directories in this repository:
This directory contains simple CMAB-based algorithms to optimize the charging of electrical batteries in a single-agent learning environment (deterministic and stochastic electricity prices) as well as in a decentralized multi-agent learning environment.
This directory contains the implementation of agents' algorithms presented in here for decentralized control of smart grids with electric vehicles.