CAREER: Distributed Coordination Strategies for Mobile Autonomous Agents

职业：移动自治代理的分布式协调策略

• 批准号: 0347285
• 负责人: Ali Jadbabaie
• 金额: --
• 依托单位: University of Pennsylvania
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2004
• 资助国家: 美国
• 起止时间: 2004-06-01 至 2010-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0347285&HistoricalAwards=false
• 关键词: CAREER; Distributed; Coordination; Strategies; Mobile

CAREER: Distributed Coordination Strategies for Mobile Autonomous AgentsDepartment of Electrical and Systems Engineering, University of PennsylvaniaThis proposal is aimed at developing tools for analysis and design of distributed algorithms for co-ordinationof groups of mobile autonomous agents. Such coordination algorithms have appeared invarious disciplines in different contexts, ranging from biological sciences (aggregation and socialcohesion in animal species) to systems and control theory and collective robotics (cooperative con-trolof multi-vehicle systems), and from statistical physics (non-equilibrium phenomena in manydegree-of-freedom dynamical systems) to computer graphics (artificial life, simulated flocking).We plan to develop a unified theory which provides a provably correct analysis and synthesis ofdistributed coordination and collective synchronization among a network of mobile agents with achanging topology. Several key problems are formulated and road-maps for obtaining efficient andscalable solutions to the problems are discussed. The proposed research plan utilizes tools fromalgebraic and spectral theory of graphs, control theory and theory of dynamical systems.Intellectual Merit: Distributed coordination problems such as collective synchronization andflocking are ubiquitous in complex biological, engineering, and physical systems. Such prob-lemsarise in control theory, physics, biology, social sciences, power system networks and in timesynchronization in sensor networks. The common feature of all of the above problems is that a setof (possibly mobile) agents are interconnected using simple nearest neighbor rules in a networkwith a topology that is time varying. We plan to develop a unified framework which allows usto study these problems and synthesize coordinated and cooperative systems of multiple agents.Moreover, this study will provide a justification for the so called emergence of collective behaviorfrom simple local interactions, a phenomenon which has been observed across complex engineer-ingand natural systems.Broader Impacts: The proposed research will impact the study of complex engineering systems,ranging from operation and joint coordination of unmanned air, ground, and underwater vehiclesas well as coordination of a network of unmanned robots to perform search and rescue operations.The research will also impact other emerging applications in homeland security, defense, trans-portation,wireless communication, and electric power networks, which require networks of agentsto be capable of multi-user communication, collaborative information processing, and distributedcomputation, actuation, and control. The research plan is also accompanied by an active educa-tionplan consisting of undergraduate and graduate level course development, which will provide aframework for studying problems involving dynamic systems over networks. The plan is designedto involve undergraduate students in research and provide them with a well-defined plan of studyto help motivate them to pursue an advanced degree. It would also bring research across severalareas of science and systems theory together in a single unifying framework.

职业：移动自主代理的分布式协调策略宾夕法尼亚大学电气与系统工程系这项提案旨在开发用于分析和设计分布式算法的工具，用于移动自主代理组的协调。这些协调算法在不同的背景下出现了不同的学科，从生物科学(动物物种中的聚集和社会结合)到系统与控制理论和集体机器人(多车辆系统的协作控制)，从统计物理(多自由度动力系统中的非平衡现象)到计算机图形学(人工生命，模拟群集)。我们计划发展一个统一的理论，为具有变化拓扑的移动代理网络之间的分布式协调和集体同步提供可证明正确的分析和综合。提出了几个关键问题，并讨论了获得这些问题的有效和可扩展解决方案的路线图。该研究方案综合运用了图论和谱论、控制论和动力系统理论等工具。智能优点：在复杂的生物、工程和物理系统中，集体同步和群集等分布式协调问题普遍存在。这些问题出现在控制理论、物理、生物、社会科学、电力系统网络和传感器网络的时间同步中。所有上述问题的共同特征是，在具有时变拓扑的网络中，一组(可能是移动的)代理使用简单的最近邻规则相互连接。我们计划开发一个统一的框架，使我们能够研究这些问题，并综合多个主体的协调和协作系统。此外，这项研究将为所谓的从简单的局部相互作用中出现集体行为提供理由，这种现象已经在复杂的工程和自然系统中观察到。广泛的影响：拟议的研究将影响复杂工程系统的研究，从无人驾驶飞机、地面和水下机器人的操作和联合协调，以及协调无人驾驶机器人网络执行搜索和救援行动。该研究还将影响其他新兴应用，如国土安全、国防、运输、无线通信和电力网络，这就要求代理网络能够进行多用户通信、协作信息处理以及分布式计算、驱动和控制。研究计划还伴随着一项积极的教育计划，包括本科生和研究生水平的课程开发，这将为研究涉及网络上动态系统的问题提供框架。该计划旨在让本科生参与研究，并为他们提供明确的学习计划，以帮助激励他们攻读更高的学位。它还将把科学和系统论的几个领域的研究集中在一个统一的框架内。