Secure control and proportional-integral-derivative performance of cyber-physical systems with sparse adversarial attacks

Abstract

This study investigates the secure control problem for discrete-time linear systems with sparse adversarial attacks. The adversarial attacker is assumed to have limited resources and the capability of manipulating a certain number of communication channels between the remote controller and the actuators. For the designer, which channels are attacked and which are not are unknown. A novel secure remote control method is established in this study. This method consists of a control law, a switching function, and a selection mechanism. The selection mechanism is designed to help select a proper feedback gain for the control law and to produce a switching function that prevents attack signals from entering the plant. Under basic and necessary assumptions, the theoretical analysis shows that the secure control problem can be transformed into a state feedback stabilization problem and that the resulting closed-loop system is stable and resilient to proportional-integral-derivative attacks. Simulation on an unmanned ground vehicle system is performed to verify the theoretical results.