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

Wenlong Liao, Birgitte Bak‐Jensen+3 more|arXiv (Cornell University)|Jan 25, 2021

Advanced Graph Neural NetworksComputer Science103 references26 citations

TL;DR

This paper surveys graph neural networks (GNNs) and their applications in power systems, summarizing GNN paradigms and key tasks such as fault analysis, time series prediction, power flow calculation, and data generation.

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 is 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 graph-structured 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 GNNs structures (e.g., graph convolutional networks) are summarized, and 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.

Motivation & Objective

• Motivate handling non-Euclidean, graph-structured data in power systems.
• Provide a comprehensive overview of classical GNN paradigms (e.g., graph convolutional networks) and their applicability to power systems.
• Review key power-system applications of GNNs, including fault analysis, time series prediction, power flow calculation, and data generation.
• Discuss main issues and emerging research trends in applying GNNs to power systems.

Proposed method

• Summarize classical GNN structures and their adaptations to graph-structured data.
• Categorize and review GNN-based applications in power systems (fault scenarios, time series, power flow, data generation).
• Highlight challenges, limitations, and potential research directions in GNNs for power systems.

Experimental results

Research questions

• RQ1 What graph neural network paradigms are most relevant for power systems?
• RQ2 How are GNNs applied to fault scenario analysis, time series prediction, power flow calculation, and data generation in power systems?
• RQ3 What are the main issues and research trends in applying GNNs to power systems?

Key findings

• The paper provides a comprehensive overview of GNNs and their structures, focusing on graph convolutional networks.
• It reviews key applications in power systems, including fault scenario analysis, time series prediction, power flow calculation, and data generation.
• It discusses main issues and research trends related to GNNs in power systems.