摘要
为了研究适用于强交互竞争型网络机器人系统的通用体系结构,提出并设计了分层的竞争型网络机器人体系结构.在该体系结构中,策略层的职责是识别对手的意图、评估威胁和生成策略.执行层的功能是克服由于观测噪声和时延等因素而造成的系统状态观测的不确定性,以及处理由于对方的敌意行动造成的己方的行动不确定性.物理接口层可将执行层的指令转化为控制机器人的底层命令.设计实现了一种高对抗度的竞争型网络机器人对弈系统——Tele-LightSaber.通过在实际Tele-LightSaber网络机器人平台上的实验,表明了该体系结构的实用性和优越性.
A hierarchical architecture is developed for the competitive networked robot system,in order to provide a general platform for competitive networked robot systems with strong interaction.In the proposed architecture,the function of strategic layer is to recognize opponent’s intention,estimate threat and generate strategy.To deal with the uncertainty in perception caused by the noise and time-delay,and that in action caused by the hostile action from the opponents,the executive layer is adopted in this architecture.The physical interface layer is used to transfer the command from executive layer into bottom command which can control robots.A typical task with high degree of opposition of competitive networked robot,named Tele-LightSaber(TLS),is proposed for the research on competitive networked robots.Finally,experimental results on TLS platform show the practicality and advantages of the proposed method.
出处
《机器人》
EI
CSCD
北大核心
2013年第4期462-469,共8页
Robot
基金
国家自然科学基金资助项目(60905061)
天津市科技计划资助项目(08JCYBJC12700)
关键词
竞争型网络机器人
分层结构
意图识别
马尔可夫决策过程
competitive networked robot
hierarchical structure
intention recognition
Markov decision process