Customized environment utilizing the OpenAI Gym library for controlling battery energy storage systems in multi-microgrids

A customized environment was created using the OpenAI Gym library to mirror battery energy storage systems in multi-microgrids. In this environment, the problem is modeled as a Markov Decision Process (MDP). This approach allows for managing complex decision-making by defining states (the various conditions of the system), actions (the available choices), and rewards (the feedback received based on the decisions). The MDP framework enables the incorporation of multiple state variables and effectively handles the dynamic interactions between different parts of the system.

This work was conducted in collaboration with the following authors:

• [email protected] (Abdelkader Messlem), [email protected] (Youcef Messlem), [email protected] (Ahmed Safa), [email protected] (Selman Kouadria), [email protected] (Djaffar Ould Abdeslam)