Event-Triggered Control Based on Neural-Network Observer for Descriptor Jump Systems Against DoS Attacks

This article studies the neural-network observer-based state feedback control problem for descriptor Markov jump systems under adaptive event-triggered mechanism and denial-of-service attacks. An improved neural-network observer is proposed, which utilizes neural network approach to approximate nonlinear function of the controlled systems, thereby achieving state estimation with accuracy. And a novel adaptive event-triggered mechanism that depends on the current sampled data, the latest transmitted data and their average value is implemented to conserve network resources. Furthermore, by establishing Lyapunov-Krasovskii functional associated with attack mode, the criteria of finite-time boundedness for closed-loop systems are given through the linear matrix inequality. Finally, the oil catalytic cracking process validates the effectiveness of the offered strategies.