Secure Cyberattack and Anomaly Detections in the Presence of Actuators Lags and Communication Delays With Application to Industrial Gas Turbines

Abstract

Cyber-Physical Systems (CPSs) utilize Networked Control Systems (NCS), where the supervised controller is connected to the network through communication links. It facilitates the network with online accessibility that reduces maintenance costs and enhances reliability. However, open access often raises the risk of cyber-attacks. Besides, a communication channel delay due to limited bandwidth is often possible, which may lead to instability problems. This paper introduces a new observer-based resilient control framework to deal with cyber-attacks and communication delays. As such, deception and Denial pf Service (DoS) actuator attacks are considered by defining appropriate unified functions. Following that, an augmented state-space model is proposed for the CPS with time delay in the communication links. Next, states are estimated by a sliding mode observer from the corrupted measurements. Then, a residual generation is appropriated to identify cyber-attacks in the augmented system and to assess the cyber-attack detection approach. The Lyapunov stability theory guarantees the convergence of the estimation. Experimental data on an industrial gas turbine is utilized to validate the system model. Test results indicate that the sliding mode observer enjoys a secure framework for network communications.