Impulse-based output consensus tracking of heterogeneous clustered networks subjected to unknown bounded cyber-attacks

This paper investigates the output consensus tracking problem of heterogeneous multi-agent systems over a complex clustered network subject to unknown sensor attacks. In the case of false data injection attacks, the attackers launch attack signals into the sensors of heterogeneous agents. Firstly, a novel network structure with inter-cluster impulsive pinning control is designed for the complex wide-area network. Secondly, by modelling sensor attacks, an adaptive dynamics is developed to estimate the unknown attack signals, then a unified framework for the adaptive observer is established using the compromised output information. Thirdly, a distributed adaptive observer-based hybrid controller under the internal model and impulsive control is designed to achieve output consensus with uniform ultimate boundedness. Furthermore, sufficient conditions of consensus tracking are developed based on the output feedback control method, where controllers and adaptive parameters are derived using both linear matrix inequalities and output regulation techniques. Finally, the developed adaptive observer-based hybrid control method is verified by the simulation results of unmanned aerial vehicle systems.