MLPER: A Machine Learning-Based Prediction Model for Building Earthquake Response Using Ambient Vibration Measurements

Deep neural networks (DNNs) have gained prominence in addressing regression problems, offering versatile architectural designs that cater to various applications. In the field of earthquake engineering, seismic response prediction is a critical area study. Simplified models such as single-degree-of-freedom (SDOF) and multi-degree-of-freedom (MDOF) systems traditionally provided valuable insights into structural behavior, known for their computational efficiency facilitating faster simulations. However, these notable limitations capturing nuanced nonlinear behavior structures spatial variability ground motions. This study focuses on leveraging ambient vibration (AV) measurements buildings, combined with (EQ) time-history data, create predictive model using network (NN) image format. The primary objective predict specific building’s accurately. training dataset consists 1197 MDOF 2D shear models, generating total 32,319 samples. To evaluate performance proposed model, termed MLPER (machine learning-based building structures’ response), several metrics are employed. These include mean absolute percentage error (MAPE) deviation angle (MDA) comparisons time domain. Additionally, we assess magnitude-squared coherence values phase differences (Δφ) frequency underscores potential reliable tool predicting responses, simplified models. By integrating AV EQ data framework, offers promising avenue enhancing our understanding during events, ultimately contributing improved resilience design engineering.