Graduation Year

2024

Document Type

Thesis

Degree

M.S.E.E.

Degree Name

MS in Electrical Engineering (M.S.E.E.)

Degree Granting Department

Electrical Engineering

Major Professor

Mia Naeini, Ph.D.

Committee Member

Ismail Uysal, Ph.D.

Committee Member

Nasir Ghani, Ph.D.

Keywords

LSTM, Machine Learning, PMU Placement, Recurrent Neural Networks, State Estimation, Time Series

Abstract

In an era increasingly marked by sophisticated cyber-attacks, this thesis investigates the critical issue of bus unobservability in smart grids and its impact on the effectiveness of cyber-attack detection and localization models. Given that unobservability is a prevalent challenge in smart grids due to various factors, researchers have developed numerous algorithms for optimal Phasor Measurement Unit (PMU) placement under scenarios of limited observability. However, these models primarily focus on enhancing network observability, often without considering whether this placement optimally facilitates attack detection. This research is driven by the hypothesis that a deeper understanding of the effects of unobservable buses can inform more effective PMU deployment strategies, thereby bolstering the grid's defenses against cyber-attacks.

The research is structured to first provide a comprehensive review of existing state estimation, detection, and localization models, emphasizing data-driven temporal analysis methods. It then delves into an in-depth experimental evaluation to assess how unobservability influences the accuracy and reliability of these models. The insights from these experiments are intended to inform utilities about the potential impacts of network unobservability on cyber-attack detection, contributing to a broader understanding that may support future PMU placement strategies.

The principal finding of this thesis is the identification of a direct correlation between the number of unobservable buses and the efficiency of attack detection and localization performance. As the count of unobservable buses escalates, there is a noticeable decline in the performance of both state estimation and detection and localization models. Accordingly, this study proposes an estimated threshold for the number of PMUs required to maintain model effectiveness before a critical decline in performance occurs. Moreover, the research delineates that certain buses exert a more significant influence on detection and localization outcomes than others, suggesting that strategic placement of PMUs at these buses can enhance detection capabilities. Additionally, this thesis evaluates the efficacy of detection and localization models under various common PMU placement strategies, concluding that, despite an increase in system observability, these strategies may not optimally support attack detection. The impact of clustered unobservability on detection models is also explored, providing insights into how it affects model performance.

In summary, this thesis provides a focused examination of how bus unobservability impacts the detection and localization of cyber-attacks in smart grids. It highlights the importance of strategic PMU placement as a critical factor in enhancing grid security. This work underscores the necessity for ongoing research in the face of evolving cyber threats, aiming to safeguard critical energy infrastructure effectively.

Download

Share

COinS