• DocumentCode
    1330385
  • Title

    Hybrid Control for Connectivity Preserving Flocking

  • Author

    Zavlanos, Michael M. ; Tanner, Herbert G. ; Jadbabaie, Ali ; Pappas, George J.

  • Author_Institution
    Dept. of Electr. & Syst. Eng., Univ. of Pennsylvania, Philadelphia, PA, USA
  • Volume
    54
  • Issue
    12
  • fYear
    2009
  • Firstpage
    2869
  • Lastpage
    2875
  • Abstract
    In this technical note, we address the combined problem of motion and network topology control in a group of mobile agents with common objective the flocking behavior of the group. Instead of assuming network connectivity, we enforce it by means of distributed topology control that decides on both deletion and creation of communication links between agents, adapting the network to the group´s spatial distribution. With this protocol ensuring network connectivity, a decentralized motion controller aligns agent velocity vectors and regulates inter-agent distances to maintain existing network links. The stability of the flocking controller is established in continuous time by means of an observability argument on a quadratic form of the graph Laplacian that exploits the time delay between link deletion and creation caused by the topology control protocol, which induces a dwell time between network switches.
  • Keywords
    algebra; graph theory; mobile agents; mobile robots; motion control; agent velocity vectors; algebraic graph theory; connectivity preserving flocking; decentralized motion controller; distributed topology control; graph Laplacian; inter-agent distances; link creation; link deletion; mobile agents; network switches; network topology control; Communication system control; Distributed control; Mobile agents; Mobile communication; Motion control; Network topology; Observability; Protocols; Stability; Velocity control; Algebraic graph theory; cooperative control; hybrid systems; multi-agent systems;
  • fLanguage
    English
  • Journal_Title
    Automatic Control, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9286
  • Type

    jour

  • DOI
    10.1109/TAC.2009.2033750
  • Filename
    5332273