项目名称: 智能机器人自主知识与自主行动推理研究
项目编号: No.61262005
项目类型: 地区科学基金项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 周生明
作者单位: 广西师范大学
项目金额: 45万元
中文摘要: 本项目着眼于"情境-应对"下智能机器人的自主知识与自主行动推理的研究。拟借鉴人类智能生成机制的转换过程"数据-信息-知识-智能",探索机器人上"机器数据-机器信息-机器知识-机器智能"的演化机理,以"机器知识"为切入点,探索机器人自主知识及自主行动推理的新理论和新方法。一方面,通过把感知器加到情境演算行动理论中及定义基于感知器的语义解释,将情境演算进行扩展,并采用"内省设计立场"建模的思想,创建机器人的自主知识库,构建既能描述机器人的行动,又能描述感知器的逻辑系统LBS。另一方面,通过扩展LBS系统的表示和推理能力,建立从外部信息到机器知识的自动转换机制及智能机器人的自主行动推理模型,研究"机器人自主知识"向"机器人自主智能"演化的基础问题,探求建立描述多任务、多情境及多信息耦合环境下独立于设计者和外部操控人员的自主行动推理的理论体系。本研究不仅具有重要的应用价值,而且具有深远的科学意义。
中文关键词: 智能机器人;自主知识;自主行动推理;知识理解;图像识别
英文摘要: The project focuses on autonomous knowledge and autonomous action reasoning of intelligent robots under the case of "situation-response". Through incorporating the transformation process of the formation mechanism of human intelligence-"data-information-knowledge-intelligence", the evolution of "machine data-machine information-machine knowledge-machine intelligence" on robots is explored, and using "knowledge machine" as the breakthrough point, some new theories and new methods for autonomous knowledge and autonomous action reasoning of intelligent robots are proposed. On the one hand, through introducing sensors into action theory in situation calculus, and defining a semantic interpretation based on sensors, the original situation calculus would be expanded, and adopting the concept of self-reflective design stance, the autonomous knowledge base would be established. In addition, the logical system LBS is also proposed. On the other hand, through extending the representation and the ability of reasoning of LBS system, the autonomous action reasoning models are established by considering the automatic conversion mechanism from external information to the knowledge of the machine. The problem of evolution on the autonomous knowledge of robots transfer to the autonomous intelligence of robots is investigated. It
英文关键词: intelligent robot;autonomous knowledge;autonomous action reasoning;knowledge understanding;image recognition