Abstract: Accurately determining the actual loads acting on lining segments is essential for ensuring the stability and safety of shield tunnel segment lining structures during construction and operation. However, due to factors such as stratigraphic distribution complexity, tunnel scale, and burial depth, existing load theories and calculation methods cannot yet provide precise estimates of external loads on tunnel structures. To address this challenge, a case study is conducted on the shield tunnel project of Bid 5 of the Beijing East Sixth Ring Road Renovation Project, and field-monitoring data are collected and analyzed. Subsequently, ANSYS finite element software is employed to numerically simulate the monitored axial force and bending moment of the shield tunnel segment lining structure. Furthermore, a genetic algorithm program is implemented in Matlab, with soil pressure and lateral pressure coefficients as inversion parameters, to perform single- and dual-parameter inversion analyses of the loads acting on super-large-diameter shield tunnels in deeply buried sandy soil strata. The research results indicate the following: (1) Significant differences in the normalized sum of squared errors for axial force and bending moment within traditional objective functions lead to substantial errors in load inversion results. To address this issue, a multi-data intelligent load inversion analysis model and an objective function construction method suitable for ultra-large-diameter shield tunnels are proposed. (2) Compared with single-parameter inversion, the introduction of the lateral pressure coefficient reduces the objective function value of the dual-parameter load inversion by 43% and improves the fit between the calculated internal forces and the measured data. (3) For a deeply buried shield tunnel with an outer diameter of 15.4 m traversing sandy soil strata, a vertical load equal to 1.74 times the soil-column pressure within the tunnel-diameter range and a lateral pressure coefficient of 0.48 are recommended.

Key words: deeply buried sandy soil strata, super-large diameter, shield tunnel, field test, genetic algorithms, load inversion