Zonotope-based attack detection for cyber-physical systems with a dynamic event-triggered scheme

Abstract

In this paper, the attack detection problem of linear discrete-time cyber-physical systems (CPSs) with unknown-but-bounded (UBB) external disturbances and measurement noise is investigated. Firstly, a dynamic event triggering scheme (DETS) utilizing zero-order holds (ZOH) to stabilize control signals is designed to enhance communication efficiency while maintaining system performance. Then, under the constructed DETS, a design criterion for the state observer is introduced to guarantee that the observer can effectively estimate the system states and achieve $l_1$ performance. Furthermore, a residual-based attack detection method is then proposed to enhance detection sensitivity and reduce the false alarm rate (FAR). Lastly, a numerical example serves to illustrate the effectiveness of the proposed approach.