Output Tracking Control of Cyber-Physical Systems Under Hybrid Attacks and Its Applications

In cyber-physical systems (CPSs) operating under complex communication environments characterized by dynamic network topologies and bandwidth constraints, the performance of control loops can be significantly degraded by various types of interference, such as signal noise, time-varying delays, and malicious attacks. This paper investigates the output tracking control (OTC) performance of CPSs under the influences of deception attack and hybrid attacks, by proposing a trade-off performance index blends tracking error and channel input power constraints. Given the stochastic nature of cyber-attacks in CPSs, the investigated deception attacks and Denial-of-Service (DoS) attacks are formally modeled as two mutually independent Bernoulli-distributed random variables. Through frequency-domain and the Youla parameterization method, this study establishes two closed-form analytical expressions. The results reveal that the OTC performance is fundamentally constrained by the inherent characteristics of the plant and the attack parameters. Finally, simulation validations are conducted through a given single-input-single-output plant and an inverted pendulum system.