The objective of this research is to investigate a reconfigurable, multilayer, intelligent agent-based control paradigm for microgrids that are smart aggregates of a strongly heterogeneous set of distributed energy resources (including distributed generation, energy storage and demand response units). The approach is to (1) identify a multilayer, multiagent-based control framework, (2) define the structure and behavioral functions of each layer, (3) develop interaction and decision-making capabilities for each agent, (4) develop adaptive reconfiguration algorithms for the multiagent based microgrid, and (5) demonstrate these concepts through multiagent system modeling and simulation studies.
Intellectual Merit: This effort applies recent advances in the fields of multiagent system theory, decision theory, game theory, computational intelligence, and machine learning to address the control issues of distributed energy resources. This research has potential to create an entirely new field in the electric power engineering discipline and revolutionize the practice of this discipline through introducing new theories, technologies, and tools. Major contributions are expected pertaining to knowledge and understanding of large-scale, efficient deployment of renewable and alternative energy resources, and enable high penetration of distributed energy resources in existing energy infrastructures.
Broader Impacts: The success of this research represents a breakthrough in building a more secure, reliable and sustainable energy infrastructure system, which will have a profound impact on every aspect of our economy and society. The tools and techniques under investigation will provide novel content for new courses, research topics, as well as hands-on learning opportunities for undergraduate and graduate students, especially for the large population of minority students, at the University of Miami. The creative aspects of this research will be used in outreach programs to cultivate students' enthusiasm in engineering.
