This thesis explores the application of Graph Neural Networks (GNNs) for forecasting net-positions in the Nordic electricity market. Two GNN architectures, GRU-GCN and FGNN, were evaluated and compared to the existing forecasting model employed in the power grid. Results demonstrate that both GNN models achieve competitive performance, highlighting their potential for leveraging the graph structure inherent in power grids. However, regional variations in forecast uncertainty and the impact of data quality and disruptions necessitate further research. This thesis contributes to the understanding of GNNs in power grid forecasting and identifies future research directions, such as developing interpretable GNN models and incorporating additional data sources, to enhance the accuracy and reliability of power grid operations.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:umu-225775 |
Date | January 2024 |
Creators | Marklund Brinell, Gustav |
Publisher | Umeå universitet, Institutionen för fysik |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0021 seconds