Scholarlink

<p>With the increasing digitization of power systems, cyberattacks such as False Data Injection (FDI) attacks pose serious threats to grid security by manipulating state estimations and triggering cascading failures. This paper presents a comprehensive framework for detecting, locating, and mitigating Load Redistribution (LR) attacks. First, we propose a novel detection algorithm based on Graph Neural Networks (GNNs) and Graph Attention Networks (GATs) that exploit the topological structure and electrical properties of smart grids. Second, after successfully identifying LR attacks, we introduce a mitigation strategy formulated as a trilevel optimization problem, in which the system operator aims to maintain secure grid operation through corrective Security-Constrained Unit Commitment (SCUC). The integrated framework is validated on the IEEE 118-bus system under various attack scenarios. Simulation results demonstrate the GNN/GAT-based detector’s ability to accurately localize attacks and the optimization-based mitigation strategy’s effectiveness in enhancing grid resilience.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Detecting and Mitigating Load Redistribution Attacks in Smart Grids with Graph Attention Networks and a Trilevel Optimization Model

  • 期刊论文

  • 发表日期：

    2026年2月12日

查看全文
  • Huda M. Abdul Abbas,
  • Maysam Kadhim

摘要

With the increasing digitization of power systems, cyberattacks such as False Data Injection (FDI) attacks pose serious threats to grid security by manipulating state estimations and triggering cascading failures. This paper presents a comprehensive framework for detecting, locating, and mitigating Load Redistribution (LR) attacks. First, we propose a novel detection algorithm based on Graph Neural Networks (GNNs) and Graph Attention Networks (GATs) that exploit the topological structure and electrical properties of smart grids. Second, after successfully identifying LR attacks, we introduce a mitigation strategy formulated as a trilevel optimization problem, in which the system operator aims to maintain secure grid operation through corrective Security-Constrained Unit Commitment (SCUC). The integrated framework is validated on the IEEE 118-bus system under various attack scenarios. Simulation results demonstrate the GNN/GAT-based detector’s ability to accurately localize attacks and the optimization-based mitigation strategy’s effectiveness in enhancing grid resilience.