Neural network-based adaptive reliable control for nonlinear Markov jump systems against actuator attacks

In this paper, the reliable control problem for a type of uncertain Markov jump systems subjected to actuator failures, malicious attacks and partially unknown transition rates (PUTRs) is under consideration. Aiming at tackling the actuator partial failures, structural uncertainty and unknown actuator attacks thoroughly, a novel adaptive neural network based sliding mode controller synthesis is developed, which confirms that the system trajectory can be moved onto the devised sliding mode surface in finite time and remain the expected performance. Then, the analysis process for stochastic stability of the desirable sliding motion with a new sufficient criterion is carried out for the closed-loop plant with uncertain PUTRs. Finally, the F-404 aircraft engine model as a simulation example of the investigated system is selected to verify the feasibility of the design method.