基于Recognition-Primed Decision模型的多智能体作战仿真(英文)  

Multi-agent System Combat Emulation Basedon Recognition-primed Decision

作　　者：孟庆操[1] 赵晓哲[2] 姜伟[2] 

机构地区：[1]西北工业大学航海学院,西安710072 [2]海军大连舰艇学院科研部,大连116018

出　　处：《系统仿真学报》2011年第2期294-299,共6页Journal of System Simulation

基　　金：Supported by the navalkey project

摘　　要：自然决策方法应用到多智能体协作决策是人工智能研究的热点问题,而作战仿真中的协作决策问题是多智能体协作决策重要的应用领域。为了建立作战仿真中的协作决策模型,将模糊集、模糊规则引入Klein提出的RPD模型中对战场环境中的不确定性信息进行处理,并建立了基于RPD模型的作战仿真多智能体体系。仿真结果表明内核为RPD模型的兵力主体能够对战场环境自主反应,并能够进行协作决策来协调统一团队的行为。It is a spotlight to do some research on Multi-agent cooperation by means of naturalistic decision-making approach.Cooperative decision-making in combat emulation is an important area in which Multi-agent cooperation will be find widely spread use.Based on RPD model by Klein,fuzzy set and fuzzy rules were utilized to represent,interpret,and compute the imprecise information from a scenario,establishing Multi-agent System architecture.The simulation results demonstrate that RPD agents can independently response to a designed combat scenario,being able to conduct cooperative decision-making in order to cooperate with teammates.

关 键 词：多智能体系统 作战仿真 自然决策方法 RPD模型 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]