Font Size: 

At present, researchers have recognized that the uncertainty of the ground motion and the soil property needs to be considered in the seismic analysis of underground structures. The interaction mechanism between soil and underground structures is very complex, and using refined numerical models for dynamic response analysis is an effective numerical simulation analysis method. However, due to the highly nonlinear characteristics of soil and structure, the calculation cost will significantly increase with the increase of random variables. In view of this, this article first establishes a finite element model and a random excitation model for the soil and subway station structure system. Subsequently, a high-dimensional space optimization point selection strategy was used to couple random ground motion variables with random soil shear wave velocity variables, and a representative point set was obtained based on the idea of minimizing Generalized F-discrepancy (GF-discrepancy). Finally, the nonlinear time history analysis (NLTHA) was performed on the representative point set. The research results indicate that the influence of random excitation on the seismic response of subway station structures cannot be ignored, and additional random shear wave velocity will increase the lateral deformation of subway station structures and reduce their dynamic reliability. This study provides a reference for seismic analysis and design of underground structures.