The modeling and monitoring of structural systems operating under dynamic loads involve addressing tasks such as simulation, system identification, and implementing maintenance/control actions. These tasks are often complicated by the evolving nature of such systems, the variability of their interactive environments, and the inherent uncertainties. Structural Health Monitoring (SHM) seeks to tackle these challenges by leveraging sensor network data to detect and localize damage. However, higher-level SHM tasks, such as anomaly detection and performance prognosis, often expose the limitations of purely data-driven approaches. This calls for integrating physics-based knowledge, grounded in underlying mechanics, to enhance SHM methodologies and support reliable digital twins. The emerging paradigm of Physics-Enhanced Machine Learning (PEML) bridges this gap by fusing physical insights with machine learning capabilities, addressing limitations inherent to purely physics- or data-based methods. This seminar explores various strategies for incorporating physics into dynamics modeling for SHM and highlights how such integration improves the interpretability of the models, which further enhances their predictive performance and robustness.

Liu Wei is a Ph.D. candidate in the Department of Industrial Systems Engineering and Management at the National University of Singapore, supervised by Prof. Tang Loon Ching and Prof. Eleni Chatzi. His research interests include structural dynamics, physics-enhanced machine learning and structural health monitoring. He obtained his Bachelor’s degree from Nanjing University in 2020.