Consensus Control of Multiagent Systems Subject to Biasing Faults: Relative Output Information-Based Approach

Di Huang,Qishao Wang,Guanghui Wen,Zhisheng Duan

DOI: https://doi.org/10.1109/msmc.2023.3345013

2024-01-01

IEEE Systems Man and Cybernetics Magazine

Abstract:In this article, relative output information is applied to tackle the consensus control problem of linear multiagent systems (MASs) under biasing faults and uncertainty in the parameters. An auxiliary system is first introduced to reconstruct the fault signal, and a class of observers is proposed to simultaneously estimate the current states and faults of the augmented MASs. According to the estimated information, distributed ${\mathcal{H}}_{\infty}$ and ${\mathcal{H}}_{2}$ consensus controllers are proposed, respectively, within this design framework. By utilizing the consensus region-based analysis technique, some sufficient criteria are derived to achieve ${\mathcal{H}}_{\infty}$ and ${\mathcal{H}}_{2}$ consensus, respectively, when a piece of spectrum information of the Laplacian matrix related to communication topology is involved. The Laplacian matrix’s spectrum can be fully decoupled from the observer and controller design process in the case of only actuator faults. Compared with the existing methods, the proposed algorithms almost exclusively depend on the relative output information with little global information of the graph, thereby making the solution more scalable and applicable to practical situations. Numerical simulations ultimately demonstrate the validity of the proposed results.