RoboCup中阵型策略的实现

机器人足球仿真比赛是检验各种多智能体系统理论的标准平台,在这个极为复杂的多智能体环境中,多个智能体需要通过协作完成共同目标,而协作可通过共享阵型获得.阵型是多智能体协作行为所要求的,它使多个智能体以有序、智能的方式进行协作.为适应RoboCup实时动态环境下多智能体间的协作需求,文中以阵型为研究对象,提出基于不同阵型转换和基于单一阵型调整的阵型策略并将其应用到机器人足球仿真比赛中,仿真结果表明结合应用这两种阵型策略提高了仿真球队的协作攻防效果.

Realization of Formation Strategy in RoboCup

RoboCup simulation match is a standard platform for researching various Multi-Agent System theories,and cooperation is required to accomplish the common goal in this complex multi-agents environment and cooperation can be obtained by sharing formation.Formation is important for agents' cooperative behavior,which enables agents to cooperate in an ordinal and intelligent pattern.For meeting the cooperation requirements under real-time and dynamic environment,this paper regards formation as the research object and proposes two formation strategies separately based on different formations transition and single formation adjustment.Simulation results indicate that with application of these formation strategies,cooperative performance of the whole team can be enhanced.