Graph Neural Networks and Their Applications in Power Systems: A Review

Graph Neural Networks (GNNs) have emerged as a powerful tool for addressing modern power system challenges by leveraging the inherent graph structure of power networks. This review systematically examines GNN applications in state estimation, stability assessment, fault detection, renewable energy integration, dynamic topology management, and cyber-physical security, where conventional methods face limitations. GNNs excel in capturing spatial and temporal dependencies while providing interpretable, physics-aligned results. The review also highlights current limitations, such as scalability and robustness against adversarial attacks, and identifies future research directions, including domain knowledge integration. This work offers a structured framework and roadmap for advancing GNN applications in power systems.