Robot Planning, Learning and Selective Perception in Stochastic Domains

9312398主任该项目将解决为移动机器人应用程序建立综合规划和学习系统方面的问题，例如跑腿、监视、目标跟踪和搜索指定类型的对象。这些应用要求机器人在不确定性、复杂性和时间压力下工作。相关控制问题的解决方案需要有效的算法来指导移动和传感以获取有关环境的信息。

9312398 Dean The project will address problems in building integrated planning and learning systems for applications in mobile robotics, such as errand running, surveillance, target following, and searching for objects of a specified type. These applications require robots to work under uncertainty, complexity, and time pressure. Solutions to the associated control problems require efficient algorithms for directing movement and sensing to acquire information about the environment.