Secure Consensus Control of Nonlinear Multiagent Systems Based on Memory Adaptive Protocol

This article investigates memory-based adaptive event-triggered (MBAET) consensus control for nonlinear multiagent systems (MASs) subject to multiple cyberattacks. Amultiattack model is formulated, incorporating deception and irregular denial-of-service (DoS) attacks, with a focus on the frequency and duration of DoS attacks. To reduce communication overhead, an improved MBAET mechanism is proposed in which a buffer at the triggering node stores historical information, allowing previously transmitted data to be exploited in the trigger design. Moreover, the designed threshold adaptation law adjusts the triggering threshold online based on stored historical data and prescribed performance requirements, to regulate the convergence process in real time. ALyapunov function incorporating communication delays and sampling instants is constructed. During DoS attack intervals, only locally available system information is utilized, which relaxes the associated functional constraints. By further adopting a piecewise Lyapunov functional, we derive stochastic mean-square consensus conditions for MASs under multiple attacks, covering both attack and nonattack intervals. The MBAET matrix parameters and controller gains are obtained using linear matrix inequality (LMI) techniques. Finally, the effectiveness and advantages of the proposed method in enhancing attack resilience and improving resource efficiency are demonstrated through a practical simulation based on an F-404 aircraft engine system.