An efficient computational model for large-scale structures based on improved hippopotamus optimization and time-domain inversion

Abstract

Seismic safety assessment of large concrete dams necessitates comprehensive consideration of soil-structure interaction (SSI) effect. However, excessive computation time for soil-structure models limits seismic samples. Furthermore, the simplified homogeneous foundation assumption neglects joints and cracks, leading to large discrepancies between simulated and measured seismic responses. To address these issues, this study proposes an efficient computational model (hereafter termed the equivalent model) construction method based on time-domain foundation model identification using the improved Hippopotamus optimization algorithm (IHO-TFMI), with the core program being open-source. Specifically, this study establishes a surrogate foundation model with clear physical mechanisms at structural boundaries, then derives a response surface between foundation model parameters and structural response errors by integrating SSI mechanisms. The IHO is proposed for the solution space features of the objective function, ultimately yielding a precise equivalent model. Three case studies with varying complexity demonstrate the method's reliability. The results show that the constructed structural equivalent model achieves excellent agreement with the SSI model, with the MSE of time-history responses reduced below 0.01. Computationally, the 3D equivalent model reduces calculation time by 92 % and storage usage by 99.8 % compared to the SSI model. Meanwhile, IHO outperforms other optimizers in global search capability, and the self-developed TFMI program reduces optimization time from days (with direct FEM software calls) to minutes. In conclusion, the proposed method provides an efficient and accurate alternative to traditional large-scale complex foundation modeling, facilitating advancements in research requiring massive seismic sample iterations, particularly in seismic fragility analysis of large-scale structures.