Leader‐Following Resilient Secure Consensus Output‐Feedback Control for Nonlinear Multi‐Agent Systems Under Replay Attacks

ABSTRACT Replay attacks represent a stealthy and destabilizing threat to multi‐agent systems and can undermine consensus performance by injecting deceptive historical data. To counteract this threat, this paper proposes a leader‐following resilient secure consensus control strategy for strict‐feedback nonlinear multi‐agent systems subject to unmeasurable states and replay attacks. First, a fuzzy state observer is constructed for each follower to approximate unknown nonlinearities and reconstruct unmeasurable system states. Second, a distributed adaptive secure controller is designed via the backstepping technique to achieve consensus objectives despite the presence of replay attacks. Crucially, a quantifiable resilience condition related to attack‐induced errors is derived through rigorous Lyapunov stability analysis. Under this condition, all closed‐loop signals are proven to remain bounded, with tracking errors converging to a small neighborhood of the origin. Finally, numerical simulations validate the effectiveness of the proposed approach.