Intelligent prediction of surrounding rock deformation of shallow buried highway tunnel and its engineering application
•Deformation is predicted by support vector machine information granulation.•Minimum, average and maximum of surrounding rock deformation are predicted.•Support vector machine parameter optimization method is optimized. The potential arch crown settlement is one of the most hazardous factors in shal...
Gespeichert in:
Veröffentlicht in: | Tunnelling and underground space technology 2019-08, Vol.90, p.1-11 |
---|---|
Hauptverfasser: | , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | •Deformation is predicted by support vector machine information granulation.•Minimum, average and maximum of surrounding rock deformation are predicted.•Support vector machine parameter optimization method is optimized.
The potential arch crown settlement is one of the most hazardous factors in shallow-buried tunnel excavations. Therefore, accurate prediction of arch crown settlement range is essential to minimize the possible risk of damage. Considering the time series regression characteristics of deformation of surrounding rock in shallow-buried tunnels, the Support Vector Machine (SVM) information granulation method was newly applied in this study for deformation prediction of surrounding rock. First, obtain monitoring data of the tunnel arch crown settlement. Second, transform the data of three arch crown settlement into a triangular fuzzy particle. The three parameters, Low, R, and Up in the fuzzy particle represent the minimum, average and maximum value of the settlement of the arch crown in three days. Then, use the SVM to predict the Low, R, and Up values of the tunnel arch crown settlement. Finally, the established prediction model of surrounding rock with SVM information granulation method was applied to the Panlongshan tunnel on the line of the Qinglan expressway in China and prediction results agree well with practical situations, which means the method of SVM information granulation used in this study could provide relatively high accuracy when applied to deformation prediction of surrounding rock in shallow-buried tunnels. Meanwhile, the SVM information granulation method is simple, feasible and easy to implement. The presented method has been validated as an effective method of deformation prediction for surrounding rock, which also has good prospects for further engineering applications. |
---|---|
ISSN: | 0886-7798 1878-4364 |
DOI: | 10.1016/j.tust.2019.04.013 |