Algorithmic design of multilevel organizational structures

A methodology for modeling and generating multilevel hierarchical distributed intelligence systems is presented. The objects that constitute such a system and the generic interactions among them are defined. A mathematical framework based on hierarchical Petri net theory, is developed for representing the interactions among these objects at the same level and across different levels. The methodology and the resulting models also provide a structured and modular way for solving the problem of designing large-scale distributed intelligence systems by breaking a computationally large problem into smaller subproblems, thus reducing the computational effort required to generate the feasible solutions. The methodology is illustrated with an example