摘要
文章首先指出了多Agent系统的理论在军事群体实时智能决策中应用的可行性,然后提出了单个实时Agent的组成结构和控制模型,讨论了在实时系统下,多个Agent间通信的不同方式和特点,通过在KQML通信原语的语义层增加时间约束,对KQML进行了扩展,解决了传统KQML通信协议无法满足实时多Agent系统的时间约束问题,实现一个基于多Agent系统的群体实时智能决策系统,为群体智能决策系统的开发提供了理论指导和实现方法。
Firstly, this paper has indicated feasibility of the application of Multi-agent system in military real-time group decision systems. Then it gives the architecture and control model of single real-time agent, discusses the varied communication way and its property, and extends KQML with timing expressions on the KQML performatives, which solve the problem of current KQML's lack of the capability to express timing constraints on actions. It also realizes a real-time intelligent group decision system, and provides guideline and method for the development of this kind of system.
出处
《系统仿真学报》
CAS
CSCD
2003年第12期1718-1720,共3页
Journal of System Simulation