Design of Hierarchically Distributed Expert Controllers for Large-Scale Systems

Hierarchical and distributed controllers have been among the most succesful control strategies for large-scale systems. However, in Control Theory most of these strategies do not utilize the functional behavior of the system. In contrast, in Artificial Intelligence strategies are developed almost completely based on the functional behavior. In this paper, a unified hierarchical control strategy is presented to utilize both mathematical and functional representations of a system. In presenting this unified strategy, an axiomatic approach is adopted. The formation of a hierarchy and a knowledge base and the distribution of control activities are all presented mathematically as structural and functional coordinability and consistency axioms.