Event-Triggered Model-Free Adaptive Predictive Control for Networked Control Systems Under Deception Attacks

The event-triggered model-free adaptive predictive control (ET-MFAPC) problem for a class of networked nonlinear control systems (NCSs) under deception attacks is addressed in this article. By using dynamic linearization technology, the NCSs are converted to an equivalent data model, and a networked MFAPC scheme with an adjustable input decay rate is constructed to compensate for the network delay. Meanwhile, the attack phenomena existing in feedback channels are modeled by considering both multiplicative and additive deception factors. Then, an ET mechanism without long-time dormancy behavior is proposed to reduce the calculation burden of the controller and save network communication resources. Rigorous convergence analysis for the proposed pure data-driven ET-MFAPC algorithm is given by employing the contraction mapping principle and it shows that the boundedness of tracking error in the mean-square sense can be guaranteed under the presented ET-MFAPC scheme. Finally, extensive simulations are performed to verify the theoretical results.