CPS: Breakthrough: Design of Network Dynamics for Strategic Team-Competition

在过去的十年里，人们对电力(智能电网)、通信、生物、社会和传感器网络等大型网络系统的认识和兴趣与日俱增。大量专注于网络系统的研究已经产生，其中主要目标一直是设计策略，通过这些策略，网络中的各个代理进行合作，以实现协调的目标。较少研究的是竞争战略场景，其中代理可能在试图实现其目标的同时进行竞争，或者可能以团队的形式竞争，使用本地通信进行本地协调。这个项目考虑了两个对立团队的竞争战略领域，其动机是应用程序，这些应用程序可以抽象地视为大量自主代理与对抗性代理或对手团队之间的竞争。一个主要的例子是控制由许多涡轮机组成的大型风电场的问题：每个旋转的叶片产生一个下游尾迹，每个涡轮机都面临着在复杂的空气动力相互作用下设置合适的转速的问题。合作控制的问题是，在邻近涡轮机的尾流效应以及难以预测的风速和风向变化的情况下，确定单个涡轮机的转速，使从风中提取的总集体能量最大化。在这个例子中，首席调查人员建议解决一个普遍的问题，即涡轮机被视为与自然竞争，不断地、对抗性地改变每个涡轮机的风速。随着分析和面向应用的研究工作的进行，将开发具有优化、数学系统理论、博弈论和集群算法等跨学科活动的教育方案。研究生和本科生都将参与，重点是吸引来自代表性不足群体的学生在整个项目期间参与研究活动。

Over the past decade there has been a growing awareness and interest in large networked systems such as those presented by power (smart-grid), communication, biological, social and sensor networks. A large body of research focused on networked systems has resulted where the primary goal has been the design of strategies by which individual agents in a network cooperate to achieve coordinated goals. Less studied are competitive-strategic scenarios where agents may be competing while trying to achieve their objectives, or may be competing in teams using local communications for local coordination purposes. This project considers the competitive-strategic domain for two opposing teams, motivated by applications that can abstractly be viewed as a competition between a large collection of autonomous agents, and an adversarial agent or team of adversaries. A primary example is the problem of controlling a large wind farm composed of numerous turbines: each rotating blade creates a downstream wake and every turbine faces the problem of setting an appropriate rotation speed under complex aerodynamic interactions. The cooperative control problem is to determine rotation speeds for the individual turbines that maximize the total collective energy extracted from the wind, under wake effects from neighboring turbines and difficult-to-predict variations in wind speeds and directions. In this example, the Principal Investigators propose to address a generalization of the problem where the turbines are viewed as competing against nature, which continually and adversarial changes the wind speed at each turbine. Ongoing with the analytical and applications-oriented research efforts will be the development of educational programs with interdisciplinary activities in optimization, mathematical systems theory, game theory and clustering algorithms. Both graduate and undergraduate students will be involved, with an emphasis on attracting students from underrepresented groups to participate in the research activities throughout the duration of the project.