RESEARCH

Scientific Machine Learning for Engineered Systems

Machine learning and deep learning are rapidly growing fields that learn representations of data via computational models, which have proven to successfully solve challenges in multiple domains. One emerging question is whether we can bridge deep learning methods with engineering and science, where models are commonly derived based on physical laws. Our group is particularly interested in developing and applying novel approaches in machine learning/deep learning for challenges in dynamical systems and engineered systems.

Relevant Research:

[discovering governing equations]

[structural damage detection 1] [structural damage detection 2]

[discrepancy modeling via Neural ODEs] [PgDMM]

Vision-based Sensing for Structural Monitoring

Sensing technologies have emerged as aided tools or solutions for continuously measuring data from systems either for the purpose of monitoring or system characterization. In this direction, the group’s research works have been exploiting the solutions for easily deployed, low costing, and data-efficient sensing.

Relevant Research:

[full-field motion tracking via event-based cameras]

[tracking continuous edge from videos]

Mobility Sensing (Ongoing)

This research investigates the possibility of shifting the sensing paradigm from a stationary manner to a mobile manner, which has the potential of increasing the spatial resolution of sensing, and of benefiting the sensing for city-scale applications.

and More is coming…

PUBLICATIONS

(For a full list of publications, go to Google Scholar)

Neural Modal ODEs: Integrating Physics-based Modeling with Neural ODEs for Modeling High Dimensional Monitored Structures
Z. Lai^*, W. Liu, X. Jian, K. Bacsa, L. Sun, E. Chatzi
Data-centric Engineering (2022)

A robust bridge weigh‐in‐motion algorithm based on regularized total least squares with axle constraints
X. Jian, Z. Lai, Y. Xia, L .Sun
Structural Control and Health Monitoring (2022)

Physics-guided Deep Markov Models for Learning Nonlinear Dynamical Systems with Uncertainty
W. Liu, Z. Lai^*, K. Bacsa, E. Chatzi
Mechanical Systems and Signal Processing (2022)

Structural identification with physics-informed neural ordinary differential equations
Z. Lai^*, C. Mylonas, S. Nagarajaiah, E. Chatzi
Journal of Sound and Vibration (2021)

Full-field structural monitoring using event cameras and physics-informed sparse identification
Z. Lai, I. Alzugaray, M. Chli, E. Chatzi
Mechanical Systems and Signal Processing (2020)

Measurement of full-field displacement time history of a vibrating continuous edge from video
S. Bhowmick, S. Nagarajaiah, Z. Lai
Mechanical Systems and Signal Processing (2020)

Sparse structural system identification method for nonlinear dynamic systems with hysteresis/inelastic behavior
Z. Lai, S. Nagarajaiah
Mechanical Systems and Signal Processing (2019)

Semi‐supervised structural linear/nonlinear damage detection and characterization using sparse identification
Z. Lai, S. Nagarajaiah
Structural Control and Health Monitoring (2019)

Adjustable template stiffness device and SDOF nonlinear frequency response
Z. Lai, T. Sun, S. Nagarajaiah
Nonlinear Dynamics (2019)

Moving-window extended Kalman filter for structural damage detection with unknown process and measurement noises
Z. Lai, Y. Lei, S. Zhu, Y. Xu, X. Zhang, S. Krishnaswamy
Measurement (2016)