Neural Network-Based Anti-Disturbance Control of Nonhomogeneous Markov Switching Systems

This brief studies delves into the exploration of anti-disturbance control for nonhomogeneous Markov switching systems subject to asynchronous phenomenon and time-varying transition probabilities. In addition to the conventional Markov process, a novel nonhomogeneous Markov process is addressed, in which the time-varying transition probabilities are influenced by an upper-level nonhomogeneous Markov process. In response to the challenge of handling unobservable disturbance information, an innovative adaptive neural network-based anti-interference control strategy with dynamic weight adjustments is devised. To accurately capture the asynchronous nature of interactions between the controller and the system, an asynchronous disturbance observer using a mode monitoring scheme is proposed. Building on the foundation of Lyapunov stability theory, sufficient conditions are carried out to ensure the bounded probability of the closed-loop system. Finally, the practicality of attained theoretical findings is verified via a practical example.