Engineering Multigroup Agents for Complex Cooperative Systems

Intelligent agents and multiagent systems (MAS) provide a scalable approach to distributed artificial intelligence [3]. Analogous to intelligent biological entities, agent-based computational systems may become quite complex. Some may consist of multiple cooperating groups, each providing a specialized subset of the total required functionality [4, 6]. My work focuses on engineering multigroup agents in cooperative intelligent systems. A reusable multigroup agent architecture has been developed, along with a simulation framework and recommended software engineering practices, to support the implementation of multigroup applications. The architecture is being used to evaluate power quality control algorithms for electrical power distribution systems (PDS).