基于MAS的航空集群体系结构探究

摘要

以多智能体理论为基础,对航空集群体系结构进行了探究.采用传统的自顶向下与基于涌现的自底向上相结合的方法,将集群系统划分为三层:系统层、任务层、自组织层;设计了三类智能体:任务决策智能体、任务管理智能体、任务执行智能体,三类智能体分别实现系统的三层功能.最后以航空集群反隐身任务为例进行分析.结果表明:该体系结构具有自适应性、鲁棒性,能够适应未来复杂多变的战争环境.%The architecture of aircraft swarms is explored based on Multi-Agent theory. First,the traditional top-down and the bottom-up method based on the emergence is proposed and the swarm system is divided into three layers which includes system layer,task layer and self-organization layer. Then,Decide Agent,Manage Agent and Execute Agent are designed and the three kinds of agents can achieve the function of system of three layers respectively. Finally,an example of aircraft swarm anti-stealth task is analyzed.And the results show that the proposed architecture possesses characteristics of adaptability and robustness,which can adapt to the complex war environment in the future.