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

Abstract

Deep neural networks have revolutionized many machine learning tasks in power systems, ranging from pattern recognition to signal processing. The data in these tasks are typically represented in Euclidean domains. Nevertheless, there is an increasing number of applications in power systems, where data are collected from non-Euclidean domains and represented as graph-structured data with high-dimensional features and interdependency among nodes. The complexity of graph-structured data has brought significant challenges to the existing deep neural networks defined in Euclidean domains. Recently, many publications generalizing deep neural networks for graphstructured data in power systems have emerged. In this paper, a comprehensive overview of graph neural networks (GNNs) in power systems is proposed. Specifically, several classical paradigms of GNN structures, e. g., graph convolutional networks, are summarized. Key applications in power systems such as fault scenario application, time-series prediction, power flow calculation, and data generation are reviewed in detail. Furthermore, main issues and some research trends about the applications of GNNs in power systems are discussed.

Large-Scale Sequential Learning for Recommender and Engineering Systems

“…As for power grids, it would be interesting to try graph neural networks for more advanced work with the network topology. Recently, graph neural networks have shown impressive results in the power systems tasks [Liao et al, 2021] due to their ability to capture dependencies in the graph-structured systems.…”

Complex-Value Spatio-temporal Graph Convolutional Neural Networks and its Applications to Electric Power Systems AI

“…Several papers have already applied real-valued GCN to power systems' data analysis and management [20]. Applications include, for example, fault localization [21], power system state estimation [22], anomaly detection [23,24], detection and localization of stealth false data injection (FDI) attacks, synthetic feeder generation [25], to name a few.…”

“…where p (k) (ξ), k = 0, • • • , K denotes the k th -order derivatives of p(ξ). Therefore, M is expressed in (20). It is obvious that M a triangular matrix and non-singular.…”

Solving AC Power Flow with Graph Neural Networks under Realistic Constraints

“…For distribution grids the focus shifts to the physical characteristics of the distribution system. A review on other applications of GNN in power systems is presented in [3]. In addition to power flow calculations, an approach to compute optimal power flows based on graph neural networks is proposed in [4].…”