Designing Decentralized Controllers for Dynamical Networks

该奖项由2009年《美国复苏和再投资法案》（公法111-5）资助。该提案的重点是动态网络的去中心化控制器的设计。 本文的目标是研究三个应用领域的设计，病毒传播控制，空中交通流量管理和通信代理算法/控制 方法将开发分散控制器的设计，使控制器利用网络拓扑结构，并克服约束和变化，在一个完全分布式的方式。智力MeritThe研究工作的重点是高性能的分散控制器的动态网络的设计。 我们提出了两种新的方法，一种是基于静态线性时不变（LTI）控制器的情况下，和其他的动态LTI控制器的情况下。在静态设计方法中将采用成本函数的优化和对偶优化的使用来找到控制器。 一个多导数和多延迟的设计方法，用于合成动态LTI分散控制器将被用于在动态的情况下。 建议的研究工作将推进控制理论领域的知识，并提供新的技术分散控制的重要问题，在存在的系统dynamics.Broader ImpactsThe建议的研究将介绍本科生控制理论的研究。 教育目标包括通过高级项目和课程开发对本科生进行研究培训。 研究成果将被纳入核心课程？网络结构动力学与控制(EE 503）在华盛顿州立大学。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).The proposal focuses on the design of decentralized controller for dynamic networks. The objective is to investigate the design for three application areas, virus spreading control, air traffic flow management, and communication-agent algorithmic/control Methods will be developed for decentralized controller design, that enable the controllers to exploit the network topology and overcome constrains and variations in a fully distributed way.Intellectual MeritThe research work focuses on the design of high-performance decentralized controller for dynamic networks. We propose two novel methods, one based on the static linear time-invariant (LTI) controller case, and the other on the dynamic LTI controller case. Optimization of a cost function and use of dual optimization will be employed in the static design approach to find a controller. A multi-derivative and multi-delay design methodology for synthesizing dynamic LTI decentralized controller will be used in the dynamic case. The proposed research work will advance the knowledge in the area of control theory, and provide new techniques for the important problem of decentralized control in the presence of system dynamics.Broader ImpactsThe proposed research will introduce the undergraduate students to control theory research. The educational objectives include research training of undergraduate students through senior projects and course development. The research results will be integrated in a core course ?Network Structure Dynamics, and Controls? (EE 503) at Washington State University.