著者
M.Alicia Perez1
タイトル
Multiagent Planning in prodigy
日時
May 1991
概要
This technical report describes an extension to the PRODIGY architecture to generate parallel plans in multiagent domains. The approach followed is a centralized one in which a single planner produces plans for several resources or agents which cooperate in achieving some goal. The algorithm presented exploits the parallelism embedded in the partial order graph that NOLIMIT, the nonlinear planner of PRODIGY, generates from a single totally ordered plan. The partial order constraints, the conflicts among operators and resources, and the synchronization necessary in operators requiring more than one agent limit the parallelism that can be achieved. An execution monitor directs the parallel execution of the agents plans. This approach is demonstrated in three different domains.
カテゴリ
CMUTR
Category: CMUTR
Institution: Department of Computer Science, Carnegie
        Mellon University
Abstract: This technical report describes an extension to the PRODIGY
        architecture  to generate parallel plans in multiagent
        domains.
        The approach followed is a centralized one in which a single
        planner produces plans for several resources or agents which
        cooperate in achieving some goal.
        The algorithm presented exploits the parallelism embedded in 
        the partial order graph that NOLIMIT, the nonlinear planner
        of PRODIGY, generates from a single totally ordered plan.
        The partial order constraints, the conflicts among operators
        and resources, and the synchronization necessary in operators
        requiring more than one agent limit the parallelism that can
        be achieved.
        An execution monitor directs the parallel execution of the
        agents plans.
        This approach is demonstrated in three different domains.
        
Number: CMU-CS-91-139
Bibtype: TechReport
Month: May
Author: M.Alicia Perez1		
Title: Multiagent Planning in prodigy
Year: 1991
Address: Pittsburgh, PA
Super: @CMUTR