Cooperative and Topology Control for Tightly Coupled Large-Scale Wireless Networked Autonomous Agents

• Applicants: Professor Dr. Hannes Frey; Professor Dr. Herbert Werner
• Subject Area: Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics ; Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: from 2016 to 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 315249503

In this project we consider cooperative control of mobile agents that interact over a wireless network. The network is set up by the nodes, i.e. no external communication infrastructure is assumed. The project goal is to develop analysis and synthesis tools that allow joint design of distributed cooperative control and communication strategies for large homogeneous groups of mobile agents which interact via a wireless network. The overall control performance expressed in terms of a suitably defined l2 performance channel depends on the local dynamic control of individual agents as well as on flow of information between agents. It is restricted by imperfections of the communication network, in particular by packet loss and time delay, as well as possibly quantization effects resulting from low bit rates. The goal is to derive conditions on the stability and the performance of multi-agent systems that depend on the decision variables of the design. These variables are the control gains, the communication network parameters and the network topology. The project will continue with our study of time-triggered distributed control of wireless networked agents. We develop tools for trading network demands against control demands in an architecture which facilitates the decoupling of stochastic aspects of wireless networks and deterministic control of agent dynamics. We study with distributed consensus and flocking two core building blocks for control of multi-agent systems. In addition to concepts well understood for time-triggered control, we will explore the potential of event-triggered control as well as event-triggered model predictive control for more efficient use of channel capacity. Our theoretical results will be substantiated by simulation studies and practical experiments with AUVs and UAVs.

• DFG Programme: Priority Programmes
• Subproject of: SPP 1914: Cyber-Physical Networking (CPN)
• International Connection: USA
• Partner Organisation: National Science Foundation (NSF)
• Cooperation Partners: Professor Yaser P. Fallah, Ph.D.; Professor Javad Mohammadpour, Ph.D.