View in EDS

Physics and fiber optic data dual-driven dynamic prediction method for dam foundation settlement.

Saved in:
Bibliographic Details
Title: Physics and fiber optic data dual-driven dynamic prediction method for dam foundation settlement.
Authors: Han, Heming1,2 (AUTHOR) hanheming@hfut.edu.cn, Zha, Fusheng1 (AUTHOR) geozha@hfut.edu.cn, Xu, Long1 (AUTHOR) xulong_2005@hfut.edu.cn, Kang, Bo1 (AUTHOR) kangbo@hfut.edu.cn, Zhang, Xuedong3 (AUTHOR) ytzxd@iwhr.com, Wei, Guangqing4 (AUTHOR) wgq@nzsensing.com, Li, Hao1,2 (AUTHOR) 2021214996@mail.hfut.edu.cn, Shi, Bin1,2 (AUTHOR) shibin@nju.edu.cn
Source: Engineering Geology. Dec2025, Vol. 359, pN.PAG-N.PAG. 1p.
Subjects: Dam design & construction, Forecasting, Land subsidence, Empirical research, Safety factor in engineering, Bayesian analysis, Prediction models, Optical fiber detectors
Abstract: Settlement monitoring and prediction are critical to the safety and stability of large rockfill dams. In this study, a physics-data dual-driven prediction framework was developed, integrating physical understanding of dam deformation mechanisms with data-driven prediction techniques. A dynamic Bayesian prediction algorithm was developed to achieve rapid settlement forecasting during construction, while sliding window optimization improved model adaptability. To accurately predict settlement behaviors during the operational period, an improved ensemble empirical mode decomposition (EEMD) combined with particle swarm optimization-support vector regression (PSO-SVR) model was proposed. Verification testing was conducted to establish an integrated on-site settlement monitoring system using fusion-spliced fiber Bragg grating (FBG) sensors. The results show that the proposed method has a settlement prediction average error of less than 11 cm during the dam construction period, while the mean absolute error and root mean square error of the predicted results during the operation period were 2.44 cm and 3.27 cm, respectively. This method can provide more accurate and reliable settlement forecasts while also providing strong support for the long-term safety management of dams. • A physics-data dual-driven prediction framework was developed • Dynamic Bayesian algorithm achieves high-precision construction-phase forecasts • EEMD-PSO-SVR model achieves reliable operational-phase settlement prediction • Fusion-spliced FBG sensors capture continuous deformation with 0.1 mm resolution [ABSTRACT FROM AUTHOR]
Copyright of Engineering Geology is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Settlement monitoring and prediction are critical to the safety and stability of large rockfill dams. In this study, a physics-data dual-driven prediction framework was developed, integrating physical understanding of dam deformation mechanisms with data-driven prediction techniques. A dynamic Bayesian prediction algorithm was developed to achieve rapid settlement forecasting during construction, while sliding window optimization improved model adaptability. To accurately predict settlement behaviors during the operational period, an improved ensemble empirical mode decomposition (EEMD) combined with particle swarm optimization-support vector regression (PSO-SVR) model was proposed. Verification testing was conducted to establish an integrated on-site settlement monitoring system using fusion-spliced fiber Bragg grating (FBG) sensors. The results show that the proposed method has a settlement prediction average error of less than 11 cm during the dam construction period, while the mean absolute error and root mean square error of the predicted results during the operation period were 2.44 cm and 3.27 cm, respectively. This method can provide more accurate and reliable settlement forecasts while also providing strong support for the long-term safety management of dams. • A physics-data dual-driven prediction framework was developed • Dynamic Bayesian algorithm achieves high-precision construction-phase forecasts • EEMD-PSO-SVR model achieves reliable operational-phase settlement prediction • Fusion-spliced FBG sensors capture continuous deformation with 0.1 mm resolution [ABSTRACT FROM AUTHOR]
ISSN:00137952
DOI:10.1016/j.enggeo.2025.108435