A spatio-temporal graph convolutional approach to real-time load forecasting in an edge-enabled distributed Internet of Smart Grids energy system

Liu, Qi; Pan, Li; Cao, Xuefei; Gan, Jixiang; Huang, Xianming; Liu, Xiaodong

Abstract

As the edge nodes of the Internet of Smart Grids (IoSG), smart sockets enable all kinds of power load data to be analyzed at the edge, which create conditions for edge calculation and real-time (RT) load forecasting. In this article, an edge-cloud computing analysis energy system is proposed to collect and analyze power load data, and a combination of graph convolutional network (GCN) with LSTM, called KGLSTM is used to achieve mid-long term mixed sequential mode RT forecasting. In the proposed edge-cloud framework, distributed intelligent sockets are regarded as edge nodes to collect, analyze and upload data to cloud services for further processing. The proposed KGLSTM network adopts a double branch structure. One branch extracts the data characteristics of mid-short term time-series data through an encoding–decoding LSTM module; the other branch extracts the data features of long term timing data through an adapted GCN. GCN is used to extract spatial correlations between different nodes. In addition, by combining a dynamic weighted loss function, the accuracy of peak forecasting is effectively improved. Finally, through various experimental indicators, this article shows that KGLSTM and weighted KGLSTM have achieved significant performance improvement over recent methods in mid-long term time-series forecasting and peak forecasting.

Citation

Liu, Q., Pan, L., Cao, X., Gan, J., Huang, X., & Liu, X. (in press). A spatio-temporal graph convolutional approach to real-time load forecasting in an edge-enabled distributed Internet of Smart Grids energy system. Concurrency and Computation: Practice and Experience, Article e8060. https://doi.org/10.1002/cpe.8060

• Journal Article Type: Article
• Acceptance Date: Feb 13, 2024
• Online Publication Date: Mar 13, 2024
• Deposit Date: Mar 18, 2024
• Publicly Available Date: Mar 14, 2025
• Print ISSN: 1532-0626
• Electronic ISSN: 1532-0634
• Publisher: Wiley
• Peer Reviewed: Peer Reviewed
• Article Number: e8060
• DOI: https://doi.org/10.1002/cpe.8060
• Keywords: Real-time Intelligent Methods, Time-series Forecasting, Graph Convolutional Network, Long Short-Term Memory, Energy Systems
• Public URL: http://researchrepository.napier.ac.uk/Output/3562957