Data-Driven Control for Continuous-Time Delayed Systems Under Denial-of-Service Attacks and Unknown Measurement Noise

Abstract

This article proposes a novel data-driven approach for studying the stabilization of continuous-time delayed systems (CDSs) with both unknown model parameters and denial-of-service (DoS) attacks. Due to the effects of DoS attacks, such controlled CDSs can be viewed as a switched system with two modes: a stable mode during dormant intervals and an unstable one during attack intervals. Because of the switching character of the CDS with DoS attacks, some well-known data-driven control methods can not be used directly. To tackle this issue, a new relationship between the dormancy and activation periods of DoS attacks is built, which offers a more reasonable description of DoS attacks. Additionally, in the presence of measurement noise, a novel lemma (Lemma 3) is introduced to establish a data-based representation of CDSs using noisy measurement data. Without the need of explicit system parameter identification, stabilization criteria and control gain matrix are derived solely based on the input-state data. Finally, the effectiveness and good performance of the proposed method are illustrated through a numerical example and a power system.