Observer‐based output feedback event‐triggered consensus tracking for linear multi‐agent systems

Xuxi Zhang,Jinbao Song

DOI: https://doi.org/10.1002/oca.3111

2024-02-19

Optimal Control Applications and Methods

Abstract:Unlike most existing works, a dual ET mechanism of observer and controller is proposed, whose triggering functions are independent of each other. In addition, the time‐varying item in the triggering functions is further extended to be a class of positive L1 functions, including some existing exponential functions as its special cases. Under the proposed DET UIO and DET control protocol, it is rigorously demonstrated that consensus tracking can be achieved asymptotically, and Zeno behavior is ruled out. This paper investigates the consensus tracking of multi‐agent systems (MASs) with general linear dynamics and directed graphs via observer‐based event‐triggered (ET) control. For each follower, a dynamic event‐triggered (DET) unknown input observer (UIO) is constructed to estimate the relative states, which utilizes the discrete relative output information among neighboring agents, and the estimate errors can exponentially converge to zero. Then, an observer‐based DET controller, which has the superiority of reducing the communication burden, is presented. Unlike most existing works, a dual ET mechanism of observer and controller is proposed, whose triggering functions are independent of each other. In addition, the time‐varying item in the triggering functions is further extended to be a class of positive L1 functions, including some existing exponential functions as its special cases. Under the proposed DET UIO and DET control protocol, it is rigorously demonstrated that consensus tracking can be achieved asymptotically, and Zeno behavior is ruled out. Finally, a numerical example is provided to verify the validity of the results.

automation & control systems,operations research & management science,mathematics, applied