Interval observer-based consensus tracking for multi-agent systems under Markovian switching topologies with stealthy attack

This article investigates the bounded consensus tracking problem for uncertain linear multi-agent systems (MASs) under Markovian switching topologies in the presence of unknown but bounded disturbances and stealthy attacks. First, a class of interval observers (IOs) is designed to estimate confidence bounds of agents states and stealthy attack signals. The stealthy attack is considered as that the output information is modified by the malicious attacker. Second, distributed control protocols under Markovian switching topologies are proposed for the MASs based on the designed IOs to achieve bounded consensus. The mean-square error bounds depend on the error bounds of disturbances. After rigorous mathematical proofs, some sufficient conditions are acquired for achieving control goals. Simulations are provided to validate the effectiveness of the proposed IOs and control protocols.