Time Sequence Machine Learning-Based Data Intrusion Detection for Smart Voltage Source Converter-Enabled Power Grid

Sadi, Mohammad Ashraf Hossain; Zhao, Dongbo; Hong, Tianqi; Ali, Mohd. Hasan

doi:10.1109/jsyst.2022.3186619

Cited by 7 publications

(1 citation statement)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The work of ref. [12] deals with Pearson correlation coefficient data points with an autoencoder/decoder and a time sequence machine learning algorithm approach to find outliers efficiently and to detect false data injection attacks (FDIAs)/bad data injection attacks (BDIAs) and DoS.…”

Section: Introductionmentioning

confidence: 99%

Cyber-Resilient Converter Control System for Doubly Fed Induction Generator-Based Wind Turbine Generators

Farrar,

Ali

2024

Electronics

Self Cite

View full text Add to dashboard Cite

As wind turbine generator systems become more common in the modern power grid, the question of how to adequately protect them from cyber criminals has become a major theme in the development of new control systems. As such, artificial intelligence (AI) and machine learning (ML) algorithms have become major contributors to preventing, detecting, and mitigating cyber-attacks in the power system. In their current state, wind turbine generator systems are woefully unprepared for a coordinated and sophisticated cyber attack. With the implementation of the internet-of-things (IoT) devices in the power control network, cyber risks have increased exponentially. The literature shows the impact analysis and exploring detection techniques for cyber attacks on the wind turbine generator systems; however, almost no work on the mitigation of the adverse effects of cyber attacks on the wind turbine control systems has been reported. To overcome these limitations, this paper proposes implementing an AI-based converter controller, i.e., a multi-agent deep deterministic policy gradient (DDPG) method that can mitigate any adverse effects that communication delays or bad data could have on a grid-connected doubly fed induction generator (DFIG)-based wind turbine generator or wind farm. The performance of the proposed DDPG controller has been compared with that of a variable proportional–integral (VPI) control-based mitigation method. The proposed technique has been simulated and validated utilizing the MATLAB/Simulink software, version R2023A, to demonstrate the effectiveness of the proposed method. Also, the performance of the proposed DDPG method is better than that of the VPI method in mitigating the adverse impacts of cyber attacks on wind generator systems, which is validated by the plots and the root mean square error table found in the results section.

show abstract

Section: Introductionmentioning

confidence: 99%