Development of Deep Learning Techniques for Predicting Dynamic Responses of Local Elements in Complex Structural Systems

In our proposed project, we aim to predict the behavior of large structures like buildings and bridges and key devices upon the large structures in response to natural hazard such as earthquakes. By combining cutting-edge deep learning techniques with the multi-fidelity concept, we’re developing a deep learning model that can predict the response time histories of not only the large-scale structure but also key devices upon the structures under various seismic loads. Furthermore, utilizing hybrid simulation equipment available at the University of Toronto, this model will be refined and validated through experimental setups that mirror real-world conditions, enhancing its accuracy and reliability. This research, a collaborative effort between University of Toronto and Seoul National University, not only pushes the boundaries of engineering and artificial intelligence but also holds promise for making our communities safer and more resilient. The expected outcome is a significant advancement in structural engineering that benefits both institutions by enhancing their research capabilities and contributing to a safer, more predictable future in facing natural disasters.

Faculty Supervisor:

Oh-Sung Kwon

Student:

Partner:

Seoul National University

Discipline:

Engineering

Sector:

Education

University:

University of Toronto

Program: