Optimal Management of a Multi-Energy Microgrid Community Based on Building-to-Grid Technology

Building-to-Grid (B2G) technology is an innovative approach that integrates buildings considered as multi-energy microgrids (MEMs) into the electrical grid, making them not only consumers, but also active participants in managing energy consumption and electricity generation. The paper proposes a model of an intelligent system “autonomous dispatcher” of an urban MEM community within the framework of the B2G concept. The structure of such a model implements learning agents based on multi-agent reinforcement learning methodology to achieve the goals of both the distribution network and the buildings connected to it within a single energy community. The proposed model relies on a multi-agent version of the PPO method to manage building demand and regulate voltage in the distribution network. The effectiveness of the proposed approach is demonstrated by the example of a real urban power district, including various types of buildings with climate control systems and distributed energy generation facilities.