Data‐driven identification and modeling of the integrated soil‐structure interaction (SSI) structure system are significantly important for real‐world structures subjected to earthquakes in various engineering disciplines. The representative mechanical modeling for the overall system with SSI effect is firstly presented, and the differential evolution algorithm is introduced for model‐based iterative identification on physical parameters including stiffness and damping coefficients. The performance and reliability of the proposed methodology are systematically investigated through a numerical model‐based parametric study. Incorporation of the model selection, the incomplete measurement and noise effect on the proposed method is revealed and discussed subsequently. Then, the present method is experimentally investigated and validated using two illustrative examples, corresponding to a 12‐story large‐scale structural model and a nine‐story real‐world building structure. The performance in terms of feasible estimation and reliable prediction by using the optimally identified physical model is demonstrated through the time history comparison of measurable acceleration responses in different seismic events.

中文翻译：

---

基于土-结构相互作用的地震激励建筑结构的数据驱动识别和建模

集成土-结构相互作用（SSI）结构系统的数据驱动识别和建模对于各个工程学科中遭受地震的真实世界结构都非常重要。首先介绍了具有SSI效应的整个系统的代表性机械建模，并引入了微分进化算法，用于基于模型的物理参数（包括刚度和阻尼系数）的迭代识别。通过基于数值模型的参数研究系统地研究了所提出方法的性能和可靠性。结合模型选择，揭示并讨论了对所提出方法的不完全测量和噪声影响。然后，本方法通过两个示例进行了实验研究和验证，分别对应于一个12层的大型结构模型和一个9层的实际建筑结构。通过对不同地震事件中可测量的加速度响应进行时程比较，证明了使用最佳识别的物理模型进行的可行估计和可靠预测方面的性能。