Dam deformation prediction model based on the multiple decomposition and denoising methods
•The iTransformer model can better achieve deformation prediction of concrete dams.•The denoising performance of EMD-SGMD-WD has strong stability.•New insight for noise reduction and prediction in dam monitoring data is provided. The presence of noise within the deformation monitor data significantl...
Gespeichert in:
Veröffentlicht in: | Measurement : journal of the International Measurement Confederation 2024-10, Vol.238, p.115268, Article 115268 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | •The iTransformer model can better achieve deformation prediction of concrete dams.•The denoising performance of EMD-SGMD-WD has strong stability.•New insight for noise reduction and prediction in dam monitoring data is provided.
The presence of noise within the deformation monitor data significantly hampers the accuracy of such analyses. This paper introduces an intelligent prediction model, which uses multiple decomposition and denoising technology, and can effectively deal with the inherent noise in the original deformation monitor data of dams and establish a more accurate deformation prediction mode The decomposition and denoising of the original deformation monitor data from the concrete arch dam were primarily accomplished using Empirical Mode Decomposition (EMD), Symplectic Geometry Mode Decomposition (SGMD), and Wavelet Denoising (WD). Subsequently, the iTransformer model was utilized to carry out the prediction and analysis of the concrete dam’s deformation. The engineering cases study demonstrate that the proposed data pre-processing technique effectively accomplishes precise noise localization and elimination. Moreover, the deformation prediction model, which is based on the iTransformer model, not only demonstrates superior predictive accuracy compared to traditional models. |
---|---|
ISSN: | 0263-2241 |
DOI: | 10.1016/j.measurement.2024.115268 |