Pursuit-Evasion Games with Complex Systems in Complex Environments

智力上的成就成功的研究计划需要在两个基本问题上取得进展。首先，必须开发新的模型和计算工具，阐明分布式感知和决策任务中可用信息的结构。其次，必须设计算法来利用这些结构来设计有效的追逐-逃避策略。这些算法为许多SAN应用提供了健壮且可证明的解决方案，例如跟踪。为了确保这些结果的实际适用性，它们将被适应于两个应用(跟踪和监视)，这两个应用可以建模为追逐-躲避游戏，并在传感器-执行器网络上实现。广泛的影响由于追逐-躲避游戏的大量应用，除了算法社区外，该项目的结果还引起许多社区的兴趣，包括控制理论、经济学和博弈论。追逐-逃避游戏的进步将有助于拉近这些社区的距离，并将导致更丰富的思想交叉授粉。为了促进这一点，这项研究的成果将作为纪录片分发。该项目的成果也直接适用于涉及传感器-执行器网络的一些自动化任务，因此，它们将通过在分布式传感和决策方面的进展参与下一代自动化技术的形成。特别是，他们的效用将在一个监测老年人的原型系统上得到演示，该系统位于纽约撒马利亚医院(特洛伊)，专门从事远程护理技术。该项目将对研究和教育的整合做出两方面的贡献：首先，将利用实验性的追逐-逃避游戏设置来吸引当地中小学的少数民族和女性学生参加新的外展计划。在伦斯勒理工学院校区附近的学校(特洛伊公立高中、特洛伊中学和第14公立学校)，有一半的学生是代表不足的少数族裔。外展项目将包括追逐-逃避游戏，学生们将与实验性的SAN设置进行较量。解决简单的追逃游戏会激发学生对科学和数学的兴趣。第二，将开发一门新的研究生课程，名为“使用传感器执行器网络的追逐-逃避游戏”。在这门课程中，学生将获得分布式传感、控制、优化和博弈论的理论背景和技能，并通过在实验装置上测试他们的算法来应用这些技能。

Intellectual MeritSuccess in the proposed research plan requires progress in two fundamental issues. First, new modelsand computational tools that shed light onto the structure of the information available in distributed sensingand decision making tasks must be developed. Second, algorithms which exploit these structures to designefficient pursuit-evasion strategies must be designed. These algorithms provide robust and provable solutions to many SAN applications such as tracking. To ensure the practical applicability of these results, they will be adapted for two applications (tracking and surveillance) which can be modeled as pursuit-evasion games and implemented on a sensor-actuator network.Broader ImpactsDue to numerous applications of pursuit-evasion games, the results of this project are of interest to anumber of communities including control theory, economics and game theory in addition to the algorithmscommunity. Progress in pursuit-evasion games will help in bringing these communities closer and will leadto a richer cross-pollination of ideas. To facilitate this, the output of this research will be disseminated as amonograph. The results of this project are also directly applicable to a number of automation tasks involvingsensor-actuator networks and therefore, they will participate in the formation of the next generation ofautomation technologies through progress in distributed sensing, and decision making. In particular, theirutility will be demonstrated on a prototype system for monitoring the elderly through a collaboration withNortheast Health, located in the Samaritan Hospital (Troy, NY) which specializes in remote care technologies.The contribution of this project to the integration of research and education will be twofold: First, theexperimental pursuit-evasion game setup will be utilized to attract minority and female students from localprimary and secondary schools to new outreach programs. Half of the student population in schools adjacentto Rensselaer Polytechnic Institute's campus (Troy Public High School, Troy Middle School and PublicSchool 14) are under-represented minorities. The outreach programs will include pursuit-evasion gameswhere the students will play against the experimental SAN setup. Solutions to simple pursuit-evasion gameswill stimulate the students' interests in science and mathematics. Second, a new graduate course titled"Pursuit-Evasion Games with Sensor Actuator Networks" will be developed. In this course, students willgain theoretical background and skills in distributed sensing, control, optimization and game theory; andapply these skills by testing their algorithms on the experimental setup.