Stabilizing nonlinear model predictive control under Denial-of-Service attack via dynamic samples selection

This article investigates the application of a nonlinear model predictive control (MPC) framework for cyber–physical system (CPS) that is transmitted over network using sample-and-hold (S/H) communication. The system is vulnerable to Denial-of-Service (DoS) attack which could disrupt the channels of communication between sensor, controller and actuator. Moreover, we design a specific robust terminal set for the S/H local controller and obtain an upper bound of the sampling intervals for this controller. Then we give the necessary conditions for sampling intervals and the amount of DoS attack that system can tolerate to ensure state convergence under attack and attack-free scenarios. To mitigate the impact of such attack and maintain stability even under adverse conditions, we propose a resilient MPC algorithm with a suitable sampling update strategy, combined with an actuator buffer for storing feasible control inputs. Numerical simulations demonstrate the effectiveness and resulting performance of the proposed framework.