Dynamic self-triggered fixed-time consensus for stochastic multiagent systems with unknown function powers under hybrid cyber attacks

This paper addresses the fixed-time consensus problem for multiagent systems (MASs) with unknown function powers under hybrid cyber attacks. Different from the common systems with known input powers, the system’s input powers studied in this paper are unknown time-varying functions. In addition, this study considers both new false data injection (NFDI) attacks and communication link attacks. The studied NFDI attacks not only result in the measured states containing unknown time-varying gain and additive terms, but also unknown time-varying powers. The communication link attacks cause the communication weights between adjacent agents to become unknown time-varying functions. By establishing new control errors, a fixed-time fuzzy consensus algorithm is developed, which compensates effects of two attacks well and ensures each agent’s output satisfies the constraint. Besides, the backlash-like hysteresis is compensated by approximating the intermediate variable. To further improves resource utilization, a dynamic self-triggered mechanism is developed in this paper, which can dynamically adjust the trigger interval and does not require real-time monitoring for errors. Based on the developed control scheme, the MASs can remain practical fixed-time stable and the Zeno phenomenon can be avoided. Lastly, the validity of the theoretical results is demonstrated by experimental simulations.