Graph neural network based hydraulic turbine data stream prediction

Variational mode decomposition

Ning Li, Jing Ren, Xin Zhou, Jun Li and Chen Xue

International Journal of Low-Carbon Technologies, 2022, vol. 17, 140-146

Abstract: As a kind of green energy with mature technology, hydropower energy is more and more widely used in our real life. As the core equipment of a hydropower station, hydraulic turbine units will experience varying degrees of vibration and aging during the process of power generation. Due to the complex internal structure and the interaction between different components, the analysis and prediction of the relevant operating data of the water turbine unit has important application value. This paper proposes a graph neural network framework for multivariate data stream prediction. In this method, a graph learning module is designed to automatically extract the one-way relationship between different components of the turbine unit. In addition, the mix-hop propagation layer and expansion layer are designed to capture the spatial and temporal correlations in hydraulic turbine data stream. Experiments show that the proposed method has higher accuracy comparing with the existing methods.

Keywords: graph neural networks; multivariate data stream; hydraulic turbine data (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://hdl.handle.net/10.1093/ijlct/ctab082 (application/pdf)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:oup:ijlctc:v:17:y:2022:i::p:140-146.

Access Statistics for this article

International Journal of Low-Carbon Technologies is currently edited by Saffa B. Riffat

More articles in International Journal of Low-Carbon Technologies from Oxford University Press
Bibliographic data for series maintained by Oxford University Press ().