Significance and formulation of ground loss in tunneling-induced settlement prediction: a data-driven study
Published in Acta Geotechnica, 2023
Recommended citation: Ren, Y., Zhang, C., Zhu, M., Chen, R., & Wang, J. (2023). Significance and formulation of ground loss in tunneling-induced settlement prediction: a data-driven study. Acta Geotechnica, 1-16. https://link.springer.com/article/10.1007/s11440-023-01859-8
Ground loss is the volume of soil excavated in excess of designed excavation. It defines the boundary conditions for ground deformation field, dominating the magnitude of tunneling-induced ground settlement. In practice, it is quantified by ground loss parameter. To date, there is no unanimously agreed formulation for ground loss parameter despite its indispensable role in developing both empirical and analytical settlement solutions. Herein, a field database is utilized to quantitatively assess existing formulations of ground loss, and to unravel its role in settlement prediction via inverse analysis. It reveals that remarkable errors can be generated by classical settlement solutions, and inversely calculated values of ground loss demonstrate strong linear correlations with R-values exceeding 0.99. This implies that the reliability of classical settlement solutions can be improved with a more accurate ground loss formulation. A data-driven formulation for ground loss is developed with aid of the random forest algorithm, it can well capture the target with an R-value equaling 0.84. The developed formulation is implemented in the O’Reilly & New solution, yielding a hybrid model for settlement. This hybrid model can well predict the settlement with an R-value of 0.84, confirming the feasibility of the data-driven formulation for ground loss.
Recommended citation: Ren, Yuhao, et al. “Significance and formulation of ground loss in tunneling-induced settlement prediction: a data-driven study.” Acta Geotechnica (2023): 1-16.