Mitigation of false data injection attacks on automatic generation control considering nonlinearities

Abdelrahman Ayad, Mohsen Khalaf, Magdy Salama, Ehab F. El-Saadany

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

As the main frequency controller in the electrical grid, the Automatic Generation Control (AGC) is one of the critical components that requires high level of security. A cyber attack on the AGC system does not only affect the operation of the AGC, but it also affects the whole system frequency. Previous research works have ignored the nonlinearities of the AGC when conducting cyber-physical threats assessment. In this paper, we first demonstrate that AGC nonlinearities cannot be ignored while studying the AGC cyber-physical security. In particular, we investigate the False Data Injection (FDI) and Time Delay (TD) attacks against a multi-area AGC systems. Second, to protect against these threats, we propose a Deep Learning (DL) approach based on Long-Short Term Memory (LSTM) framework. The two-stage LSTM framework first detects data anomalies from FDI and TD attacks and then mitigates the compromised signals in order to minimize the attacks effect on the system. We evaluate and quantify the performance of our model using a two-areas AGC system. The results confirm the accuracy of the detection model and its ability to detect and determine the compromise signals. Additionally, the mitigation model significantly reduces the attacks effect on the AGC system.

Original languageBritish English
Article number107958
JournalElectric Power Systems Research
Volume209
DOIs
StatePublished - Aug 2022

Keywords

  • Automatic generation control
  • Cyber-physical security
  • Deep learning

Fingerprint

Dive into the research topics of 'Mitigation of false data injection attacks on automatic generation control considering nonlinearities'. Together they form a unique fingerprint.

Cite this