At the plant level, a smart grid unifies power grids, consumers and generating
facilities in a single automated system which enables real-time tracking and
monitoring the regimes of all participants. It responds automatically to the changes
of all parameters in the power grid and maintains no-break power with maximum
benefits and lower human involvement. The most effective control can be attained in
smart grids by using intelligent agents. Their functionalities should be based on
intelligent control algorithms with predictive models featuring high-precision
treatment of process knowledge and adaptive learning. This chapter offers a concept
of developing an intelligent multi-agent system that maintains stability of Russia’s
Smart Grid incorporating an active analytical network (AAN) and new algorithms to
determine the degree of the system’s stability by using Gramians.