Switching formation strategy with the directed dynamic topology for collision avoidance of a multi-robot system in uncertain environments

Vu Phi Tran*, Matthew A. Garratt, Ian R. Petersen

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    23 Citations (Scopus)

    Abstract

    This paper tackles the distributed leader-follower cooperative control problem for networked heterogeneous unmanned aerial vehicle-unmanned ground vehicle (UAV-UGV) systems in unknown environments requiring formation keeping, obstacle avoidance, inter-robot collision avoidance, and reliable robot communications. To adopt various formations, we design a novel negative imaginary (NI) switching formation protocol with a directed dynamic topology. To prevent an inter-mobile robot collision, a new method to formulate the virtual propulsive force between robots is employed. To avoid unexpected obstacles, a new obstacle avoidance technique that allows the UGVs' formation to change its shape and the UGVs' roles is developed. To determine each UGV robot's order in obstacle avoidance formation, a quadrotor UAV, controlled by a strictly negative imaginary controller involving good wind resistance characteristics, tracks the center of formation shape to guarantee the maintaining visibility for multi-robot systems on the ground. The proposed control system's efficacy is investigated through a rigorously comparative study with other control techniques, namely, the performance of artificial potential field and an NI obstacle avoidance strategy using the switching formation control method without switching topology. Finally, we also conduct a stability analysis of the closed-loop control system using the NI-systems theory.

    Original languageEnglish
    Pages (from-to)2948-2959
    Number of pages12
    JournalIET Control Theory and Applications
    Volume14
    Issue number18
    DOIs
    Publication statusPublished - 17 Dec 2020

    Fingerprint

    Dive into the research topics of 'Switching formation strategy with the directed dynamic topology for collision avoidance of a multi-robot system in uncertain environments'. Together they form a unique fingerprint.

    Cite this