Active Power Management in Multiple Microgrids Using a Multi-Agent System with JADE

Abstract

This paper presents a Multi-Agent System (MAS) to model and enable an active power management in a multiple microgrids system consisting of batteries, photovoltaic and diesel micro sources in islanded or grid-connected operation. The multi-agent system was developed in JADE (Java Agent DEvelopment Framework), a Foundation for Intelligent Physical Agents (FIPA) and a compliant open source multi-agent platform. All the six case studies presented in this paper were modeled using JADE. The cases were chosen considering the real life possible scenarios, such as: discharging the battery (boost mode), energy export from the microgrid, charging the battery (buck mode), energy import from diesel generator, energy import from the grid and the system halt. However, considering that the results for the first two cases can demonstrate the performance and features of the proposed MAS, this paper focuses on the results of these cases. The results demonstrate the effectiveness of the proposed control algorithms, necessary for coordinated control and power management, and show the possibility of autonomous built-in operation of a microgrid with a multi-agent system using JADE.