Deferred actuator replacement-based performance recovery consensus control for nonlinear multi-agent systems under DoS attacks
摘要
This paper investigates consensus control for nonlinear multi-agent systems with couplings, Denial-of-Service (DoS) attacks and dynamic actuator redundancy under limited communication bandwidth. In such settings, the simultaneous presence of aperiodic transmissions, DoS-induced information loss, and deferred actuator replacement makes it challenging to guarantee both tracking performance and post-fault recovery. To address these issues, a communication-efficient fault-tolerant consensus scheme is developed by combining a dynamic self-triggered mechanism, so that transmissions are validly regulated. A distributed resilient formation filter is designed to reconstruct the leader state under DoS attacks, while a prescribed-performance control framework is embedded into the consensus design so that transient and steady-state bounds of all tracking errors are explicitly enforced even in the presence of actuator faults. Furthermore, a reconfigurable controller is constructed around a dynamic redundancy architecture with deferred actuator replacement, where a shifting function reshapes the tracking error at the switching instant and yields explicit relationships among detection delay, warm-up duration, and post-switch settling time, leading to a certified upper bound on the recovery time. It is proved that all closed-loop signals remain bounded and that tracking errors converge asymptotically, and simulations demonstrate fast post-fault recovery and significant communication savings under DoS attacks and actuator switching.