Edge-Based Dynamic Event-Triggered Leader-Following Output Consensus of Uncertain Heterogeneous Multi-Agent Systems With Multi-Channel Hybrid Cyber-Attacks
Xinchun Jia, Hongpeng Li, Xiaobo Chi, Teng Lv
Abstract
This article investigates the leader-following output consensus (LFOC) of the uncertain heterogeneous multi-agent systems (MASs) under a multirate sampling (MRS) mechanism and multi-channel hybrid (MCH) cyber-attacks by using an internal model principle. A novel form of cyber-attacks called MCH cyber-attacks is introduced, which encompasses distributed Denial of Service (DDoS) attacks and distributed deception (DD) attacks. First, we devise an MRS mechanism to sample real-time outputs from each agent with multiple physical properties. A multirate observer for each agent to fuse the sampled-data at different rates while reconstructing the agents’ states. Second, we develop an edge-based dynamic event-triggered mechanism (DETM) with flexibility and high scalability to alleviate the communication burden. The triggering conditions of this mechanism can be adjusted in real-time based on the requirements of each communication channel. With the multirate observer and edge-based DETM, we construct a fully distributed observer-based controller for each agent. Under this controller, the LFOC problem of uncertain heterogeneous MASs subject to the MCH cyber-attacks can be solved in a fully distributed manner, while ensuring the exclusion of the Zeno behaviors. Finally, a simulation example is presented to validate the effectiveness of the proposed method.