Title: Planning and Uncertainty: On the Logic of Multi-Agent Plans and their Entropy

Author: Eric Werner, Ph.D.


Contact: eric.werner@cellnomica.com

Abstract

The logical foundations of plans in the context of uncertain knowledge about the state of the world and uncertain knowledge about the plans of other agents are investigated. The partial plans of single and groups of agents are formally defined and their properties are studied. The concept of entropy is extended through a mathematical definition of the control entropy of plan states. Furthermore, fuzzy plan states are defined. A tensed modal language, containing can and plan operators, is given a formal semantics. The relevance to communication and cooperation among multiple agents in distributed AI is discussed.