Abstract Title

Machine learning based surrogate modeling with SVD enabled training for nonlinear dynamic systems

Author Information

Megan ButcherFollow

Is this project an undergraduate, graduate, or faculty project?

Undergraduate

individual

Daytona Beach

Authors' Class Standing

Megan Butcher, Junior

Lead Presenter's Name

Megan Butcher

Lead Presenter's College

DB College of Engineering

Faculty Mentor Name

Siddharth Shiladitya Parida

Abstract

Performance-based earthquake engineering is a comprehensive decision-making tool used in the computationally expensive estimation of engineering demand parameters (EDPs) via finite element (FE) models. Earthquake and parameter uncertainty limits the use of the Performance-Based Earthquake Engineering framework therefore, attempts have been made to substitute FE models with surrogate models, however, these models are a function of building parameters only. This necessitates re-training for earthquakes not previously seen by the surrogate. This research proposes a machine learning-based surrogate model framework, which considers both these uncertainties in order to predict unseen earthquakes. Earthquakes are characterized by their projections on an orthonormal basis, computed using the SVD of a representative ground motion suite. This generates a large variety of earthquakes by randomly sampling weights and multiplying them with the basis. The weights along with the constitutive parameters serve as inputs to a machine learning model with EDPs as the desired output. Four competing machine learning models were tested and it was observed that a deep neural network (DNN) gave the most accurate prediction. The framework is validated by using it to successfully predict the peak response of one-story and three-story buildings represented using stick models, subjected to unseen ground motions.

Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, or Ignite Grants) from the Office of Undergraduate Research?

Yes, Student Internal Grants

Share

COinS
 

Machine learning based surrogate modeling with SVD enabled training for nonlinear dynamic systems

Performance-based earthquake engineering is a comprehensive decision-making tool used in the computationally expensive estimation of engineering demand parameters (EDPs) via finite element (FE) models. Earthquake and parameter uncertainty limits the use of the Performance-Based Earthquake Engineering framework therefore, attempts have been made to substitute FE models with surrogate models, however, these models are a function of building parameters only. This necessitates re-training for earthquakes not previously seen by the surrogate. This research proposes a machine learning-based surrogate model framework, which considers both these uncertainties in order to predict unseen earthquakes. Earthquakes are characterized by their projections on an orthonormal basis, computed using the SVD of a representative ground motion suite. This generates a large variety of earthquakes by randomly sampling weights and multiplying them with the basis. The weights along with the constitutive parameters serve as inputs to a machine learning model with EDPs as the desired output. Four competing machine learning models were tested and it was observed that a deep neural network (DNN) gave the most accurate prediction. The framework is validated by using it to successfully predict the peak response of one-story and three-story buildings represented using stick models, subjected to unseen ground motions.