A Multi-Agent Based Approach for Energy Management in Microgrids

A shift to a decentralized power generation mainly based on Renewable Energy Sources (RESs) is essential to meet the climate protection goals of reducing greenhouse gas emissions. The fluctuating generation of RESs poses challenges for the stability of the electrical grid. Therefore, energy management systems (EMSs) are essential to cope with the volatility of RESs through an intelligent coordination of Distributed Energy Resources (DERs). Thus, EMSs ensure the security of energy supply and enhance the integration of RES. In this work, an energy management strategy based on a multi-agent system (MAS) for a microgrid is presented. The control strategy exploits the flexible operation of the DERs deployed in heating systems, which arises from using thermal storages, to adapt to the RESs fluctuations. This is performed by determining optimized schedules for the DERs using Mixed Integer Linear Programing (MILP) as well as thermal and electrical demand forecasting models. The coordination of the DER within the microgrid is achieved through a centralized and a decentralized approach for determining the operation schedules. Both approaches are assessed against a scenario in which the conventional heat driven operation of DERs is employed. The evaluation criteria comprise Primary Energy Consumption (PEC), reduction of CO2 emissions, operation costs and integration of RES. Both centralized and decentralized approaches achieve significant reduction in PEC and CO2 emissions compared to the heat driven scenario. The centralized approach achieves the highest coordination level by minimizing the import of external electricity and the highest integration of RESs. The decentralized approach is characterized by a high electricity export and achieves the lowest PEC and lowest net CO2 balance.