Data-Driven Dynamic Event-Triggered Control for NCSs Under Denial-of-Service and Time-Varying Delays

This article studies a data-driven co-designing method for networked control systems under denial-of-service attacks and communication delays, using only noisy data of system trajectories instead of explicit system models. To this end, a delay system framework is first established by investigating the effect of both denial-of-service attacks and time-varying delays. In order to reduce the consumption of the network resource, a novel dynamic event-triggering scheme is proposed. The sensor is time-triggered while the controller and the actuator are event-triggered in the scheme. Based on the Lyapunov stability approach and a data-based representation of unknown systems, a data-driven method for co-designing the controller gain and event-triggering parameters is developed. In the presence of denial-of-service attacks with unknown patterns, the proposed data-driven co-designing method demonstrates the trade-off between the system performance metric and the denial-of-service resilience level. Furthermore, the impact of time-varying communication delays on the denial-of-service resilience level is investigated. Finally, several numerical experiments demonstrate the effectiveness of the proposed methods.