CAREER: Coordination of Dynamic Networks - A Hybrid System Approach

Oklahoma State University, Stillwater OK

Investigators

Abstract

This Career development plan includes significant research, educational, and outreach components. The research component is focused on developing novel and systematic methodologies for the analysis, design, and implementation of cooperative dynamic networks. A dynamic network consists of spatially distributed dynamic nodes (e.g., autonomous vehicles, mobile sensors) which are coordinated by common set of goals and possible dynamic interaction between the nodes. There are many applications where a dynamic network may be more suitable than a single vehicle, especially where a distributed system of sensors is advantageous. Yet without coordinating the movement of the nodes or without managing the information the nodes retrieve, any advantage of dynamic network deployment may be lost and damaging collisions or interference may occur. This research will be integrated into curriculum development designed to provide research opportunities for undergraduate students with emphasis on attracting and retaining underrepresented minorities. The outreach component will support on-going educational programs designed to target K-12 institutions, and the Native American populations of Oklahoma. Intellectual Merit New Developments in complex networks of interacting systems like teams of au- tonomous vehicles/robots for homeland security, search and rescue operations, disaster relief operations, multi-targeting/multi-platform battlefield groups, unmanned air vehicle systems, intelligent highway/vehicle systems, wireless surveillance networks, and elsewhere place severe demands on the design of cooperative control schemes and communication strategies in the presence of changing environment. The large-scale in- terconnected nature of such systems requires logic-based supervisors with increased capabilities for dynamic planning, optimality and that guarantees performance subject to constraints, con ict resolution, failure handling, and other decisions. To facilitate the design of this class of systems the following key research challenges will be addressed: Development of computationally efficient algorithms for distributed model predictive control. Derivation of novel algorithms for cooperative optimization based control. Development new methodologies that enable goal directed sensing by fusing the information from heterogeneous sensors over the dynamic network, and providing the required information for each node to plan and control its mobility. Designing adaptive robust controllers that take into account the uncertainties on the dynamics of the robotic network. Experimental verification and validation of the proposed methodologies. Broader Impact Ensuring that the above Career research thrusts have the broadest impact on society, they should be integrated with a coherent educational and outreach plan. It is well-known that robotics pro- vides an excellent platform for getting students interested in science and engineering. This proposal includes educational components targeted at high school, undergraduate, and graduate students. The proposed work includes: Mentoring activities that include under-represented students in Science, Mathematics, Engineering and Technology (SMET) with the objective of retaining them in electrical and computer engineering. A robotic extension program will be designed to stimulate the interest of Native American children to increase their appreciation and understanding of science, technology, and engineering, in an informal learning environment. Seek industrial/governmental partnerships to transfer technology. Besides publications in refereed conference and journals, new courses such as Computer Embedded Systems, Robotics and Perception, and Control Hybrid Systems will be introduced into the curriculum. Software tools, algorithms, demos and results will be available on the web. The educational and research components of this Career proposal build upon previous and on-going work and existing equipment supported by Oklahoma State University.