Abstract
Microgrid systems are built to integrate a generation mix of solar and wind renewable energy resources that are generally intermittent in nature. This paper presents a novel decentralized multi-agent system to securely operate microgrids in real-time while maintaining generation, load balance. Agents provide a normal operation in a grid-connected mode and a contingency operation in an islanded mode for fault handling. Fault handling is especially critical in microgrid operation to simulate possible contingencies and microgrid outages in a real-world scenario. A robust agent design has been implemented using MATLAB-Simulink and Java Agent Development Framework technologies to simulate microgrids with load management and distributed generators control. The microgrid of the German Jordanian University has been used for simulation for Summer and Winter photovoltaic generation and load profiles. The results show agent capabilities to operate microgrid in real-time and its ability to coordinate and adjust generation and load. In a simulated fault incident, agents coordinate and adjust to a normal operation in 0.012 seconds, a negligible time for microgrid restoration. This clearly shows that the multi-agent system is a viable solution to operate MG in real-time.
Highlights
Multi-Agent Systems (MAS) have emerged as appealing technologies for Microgrid (MG) real-time operation and control
The German Jordanian University (GJU) MG consists of five Photovoltaic (PV) array sites with a total capacity of 2.11 MW, six backup diesel generators with 150 703 KVA, three load busses with campus buildings interconnected in a ring structure that integrates with the power grid through T5 point of common coupling (PCC)
A novel decentralized multi-agent system is presented for MG operations in both grid-connected and islanded modes
Summary
Multi-Agent Systems (MAS) have emerged as appealing technologies for Microgrid (MG) real-time operation and control. Research efforts have been made on the design and implementation of MAS control strategies in the context of energy management. Such implementation is still considered in the early stages of development. The main contribution of this paper is to provide a viable MAS approach with agent technical design that accounts for temporal and spatial operations of each MG renewable resource and load elements while maintaining generation-load balance in real-time and reliability and contingency constraints including optimal power flow, voltage limits, and spinning reserve. MATLAB-Simulink and Java Agent Development Framework (JADE) technologies are used for agent design to operate microgrids with load management and distributed generators control in both islanded and grid-connected modes.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
