一种基于行为的双层动态智能体结构及其在RoboCup中的应用

该文针对目前智能体结构模型的问题,提出了一种基于行为的双层动态智能体结构模型,它是一种混合结构模型,包括有反应式结构和慎思结构,并采用自信度来连接这两种结构,既可以提高在实时动态环境下智能体反应的敏捷性,也使自主机器人能够在动态环境下识别任务。这种混合结构模式已成功地应用于RoboCup仿真机器人足球赛中,并取得了比较好的成绩。

A Behavior-based Dual Dynamic Agent Architecture and its Application in RoboCup

This paper proposes a behavior-based dual dynamic agent architecture model aimed at the problems of con-ventional agent architecture models.It is a hybrid architecture model,which includes the reactive and deliberative archi-tectures,in which uses the agent's confidence to link up the two types of different architectures.The proposed architec-ture model can improve the agent's delicacy under real-time dynamic environment ,and enable the agent to identify tasks under dynamic environment.This hybrid architecture has been applied,with success,in the RoboCup simulation game ,and achieved a preferable result.