Fully Distributed Control of Microgrids Using Multi-Agent Approach

In an islanded microgrid, the main objective of a distribution operator is to manage the generation load balance based on the available resources. With the rapid integration of renewable energy and low inertia-based resources in today's power grid, it is becoming challenging to achieve optimal operation by just using conventional control techniques. This work aims to create a fully distributed control system that can efficiently manage the functioning of a microgrid (MG). The MG system should be able to maintain nominal frequency and voltage as well as synchronize between the inverters and power grid autonomously. To achieve this, we implement a hybrid control architecture employing distributed implementation with a hierarchy via a Multi-Agent System (MAS). Additionally, we propose a secondary control for frequency restoration and a control algorithm for autonomous load connection/disconnection. To develop the microgrid model under study MATLAB Simulink was used, and the control architecture was validated using several test cases through controller-in-the-loop hardware testing (C-HIL) using Raspberry Pi as an external controller.