Energy community management system based on real-time measurements and genetic algorithms

In this work we study a real-time energy management system (EMS) for Energy Communities (EC). A mathematical model of EC and the related energy optimization problem have been developed. To ensure the usability of the studied EMS in real world - where EC are varying in size and types of loads/energy generators/storage systems - we employ genetic algorithms (GA) for sub-optimal solution searching of the optimization problem that uses real-time measurements of EC production, loads and storages state of charge. We also suggest a reinforcement learning (RL) approach to enhance the performance of the EMS developed.