hide
Free keywords:
-
Abstract:
This paper studies the bipartite mean-square bounded synchronization of coupled neural networks (NNs) under anti-attack aperiodic intermittent control (AIC). A deception attack model targeting controller–actuator channels in antagonistically coupled NNs is proposed, addressing integrity breaches in communication channels and malicious command injections via intermittent access points. By incorporating averaging into intermittent control, the proposed strategy substantially enhances synchronization robustness of antagonistic networks against deceptive actuators. Through rigorous analysis employing the average AIC interval methodology, some sufficient conditions ensuring bipartite mean-square bounded synchronization for the coupled NNs are established, and the traditional strict upper/lower bounds on AIC width parameters are relaxed. To ensure the synchronization errors remain within the prescribed upper bound, the coupling strength, attack probability and the average AIC width are co-designed. Elastic interval boundary conditions for aperiodic control are derived via an average control duration analysis. Finally, numerical examples are given to demonstrate the derived results.
