Resilient Control Design for Load Frequency Control System Under False Data Injection Attacks

Abstract

Smart power grids are being enhanced by adding a communication infrastructure to improve their reliability, sustainability, and efficiency. Despite all of these significant advantages, their open communication architecture and connectivity renders the power systems’ vulnerability to a range of cyberattacks. This article proposes a novel resilient control system for load frequency control (LFC) system under false data injection (FDI) attacks. It is common to use encryption in data transfer links as the first layer of defending mechanism; here, we propose a second defense layer that can jointly detect and mitigate FDI attacks on power systems. In this article, we propose a new anomaly detection technique that consists of a Luenberger observer and an artificial neural network (ANN). Since FDI attacks can happen rapidly, the observer structure is enhanced by the extended Kalman filter to improve the ANN ability for online detection and estimation. The resilient controller is designed based on the attack estimation, which can eliminate the need for control reconfiguration. The resiliency of the proposed design against FDI attacks is tested on the LFC system. The simulation results clearly show that the proposed control system can successfully detect anomalies and compensate for their adverse effects.