智能层架构模式与仿真机器人系统

An Intelligence Layer Learning Model and Simulated Robotic Soccer Systems

以RoboCup仿真组为平台,提出一种有三层智能层的架构模式,该模式依据智能水平高低,将多智能体系统划分为被动智能层、主动智能层与社会智能层。该模式已应用于南京邮电大学Apollo2D足球机器人的比赛程序中,其有效性通过RoboCup仿真球赛得到验证。

This paper takes RoboCup Simulation as a platform to propose a novel three-layer learning model. According to levels of intelligence, Multi-agent System （MAS） in this model consists of a passive intelligence layer, an active intelligence layer and a social intelligence layer. This model has been applied to the practical competition programs of Apollo2D of the university, and its effectivity has been verified by RoboCup competitions.