Resilient control of cyber-physical systems under sensor and actuator attacks driven by adaptive sliding mode observer

The problem of resilient control of linear cyber-physical systems with cyber-attacked sensor measurements and actuator commands is studied in this article. Online reconstruction of unknown cyber-attacks is accomplished by an adaptive sliding mode observer with a novel injection term in the form of a dynamically extended equivalent control. Then, the attacked/corrupted sensor's measurements and the actuators are proposed to be cleaned up using the reconstructed attacks. The performance of the system prior to the attacks is shown to be retained after a transient response required for the attack reconstruction/estimation. The efficacy of the proposed algorithms is validated on an electrical power network.