摘要
针对Robocup自主机器人比赛中,机器人进攻和防守状态、动作转换延迟问题,采用一种基于有穷自动机模型的自然语言状态转换方法,把机器人、环境等自然语言信息抽象成形式化的符号和状态,并转化为有穷自动机输入符号.通过设定状态转移函数、字母表规则、无差别地完成各种状态转换过程.研究结果表明:该方法能有效地完成自然语言状态下的机器人状态、动作间的转换,具有较好的实时性和可靠性.
This paper proposes an action transition approach for natural language based on finite automata model to overcome the delayed action transition in offense and defense transition in Robocup for autonomous mobile robots game.In this model,the natural language information such as robots and environment is transformed into the abstract symbols,states and input symbols of finite automata.The state transition is completed by setting state transition functions,alphabet rules without any differentiation.The experimental study proves that this approach is able to efficiently complete the state and action transitions under natural language state.It is highly reliable and in real-time.
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2012年第6期885-888,共4页
Journal of Liaoning Technical University (Natural Science)
基金
国家自然科学基金资助项目(60874002)
