Numerical Simulation and Machine Learning for Pipeline Settlement Prediction in Urban Metro Station Construction: A Case Study of Shenyang Metro
Infrastructures · 03.11.2026 · 11 (11) · 403
DOI: 10.3390/infrastructures11110403 ↗Kurz gefasst
Fachartikel in der Zeitschrift Infrastructures, online erschienen am 03.11.2026. Verfasst von Ling Zhou und 6 weiteren Autor:innen. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.
Themen: Künstliche Intelligenz in der Pflege, Simulation in der Pflegeausbildung
Abstract (Originalsprache)
This study uses a subway station in Shenyang as a case and employs FLAC3D software to investigate the deformation patterns of underground pipelines during PBA construction, analyzing the influence of parameters such as burial depth, pipe diameter, elastic modulus, and soil conditions on pipeline deformation. Contour analysis of settlement indicates that pipeline settlement is positively correlated with burial depth but negatively correlated with pipe diameter, pipe material elastic modulus, and soil elastic modulus. The grey relational sensitivity analysis reveals that the main factors affecting maximum pipeline settlement, ranked by their degree of correlation from highest to lowest, are: burial depth, pipe diameter, soil elastic modulus, pipe material elastic modulus, and distance from the station axis. Based on simulation data from PBA construction, combined with data augmentation techniques, a pipeline deformation database was established and expanded. The predictive performance of eight models—SMOTER-KOA-LSB, SMOTER-KOA-GAM, KOA-GAM, KOA-LSB, GAM, LSB, RF, and SVM—was compared, and results show that the SMOTER-KOA-LSB model performs best in predicting pipeline deformation. Addressing the complex interactions among stations, soil, and pipelines caused by underground pipeline deformation during subway station construction, this study uses numerical simulation methods to explore these issues and achieves meaningful research outcomes.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Zhou, L., Hu, R., Li, X., Chen, X., Guo, Z., Yan, P., & Sun, J. (2026). Numerical Simulation and Machine Learning for Pipeline Settlement Prediction in Urban Metro Station Construction: A Case Study of Shenyang Metro. Infrastructures, 11(11), 403. https://doi.org/10.3390/infrastructures11110403
Vancouver
Zhou L, Hu R, Li X, Chen X, Guo Z, Yan P, et al. Numerical Simulation and Machine Learning for Pipeline Settlement Prediction in Urban Metro Station Construction: A Case Study of Shenyang Metro. Infrastructures. 2026;11(11):403. doi:10.3390/infrastructures11110403
BibTeX
@article{Zhou2026,
author = {Zhou, Ling and Hu, Ruzhen and Li, Xiaojing and Chen, Xiangzhe and Guo, Zhoupeng and Yan, Peng and Sun, Jingchao},
title = {Numerical Simulation and Machine Learning for Pipeline Settlement Prediction in Urban Metro Station Construction: A Case Study of Shenyang Metro},
journal = {Infrastructures},
year = {2026},
volume = {11},
number = {11},
pages = {403},
doi = {10.3390/infrastructures11110403}
}
EUVIMED