Model-free adaptive control for nonlinear multiagent systems with hybrid cyberattacks and unknown disturbance

This paper proposes a novel model-free adaptive control (MFAC) scheme for nonlinear discrete-time multi-agent systems facing hybrid cyberattacks and unknown disturbances. The proposed scheme has two main features. First, it integrates a compensation mechanism to handle aperiodic Denial-of-Service (DoS) attacks and is robust against persistent deception attacks, relying solely on real-time input/output data. Second, a Radial Basis Function Neural Network (RBFNN) is employed to estimate and counteract unknown disturbances online, which significantly enhances system robustness. Theoretical analysis confirms that the tracking errors remain bounded under the proposed framework.