Seeking the Intelligent Data-Physics Integration for Complex Engineered Systems
Electro-Chemo-Mechanics | Energy Storage Systems | Scientific Machine Learning
Alert: Our systems are becoming increasingly complex!
Five out of the fourteen Grand Challenges for engineering in the 21st century identified by the National Academy of Engineering lie in the difficulty of deciphering complex natural or engineered systems. An extremely large example is the earth’s climate, the understanding of which this year’s Nobel Prize in Physics was awarded for. A smaller example is energy storage systems (ESSs) such as lithium-ion batteries (LIBs), the main technology that the 2019 Nobel Prize in Chemistry was awarded for. The complexity of such systems comes from two major aspects: i) multiple length- and time-scales and ii) Coupled physical effects, such as mechanics, chemical reactions, and mass transport in LIBs.
What is Data-Physics Integration?
The fundamental challenge of systematic engineering lies in the tradeoff between the abundance of data and the adequacy of physical laws. At the microscale, physics can usually be elucidated, but data are expensive and limited; at the macroscale, physical laws are often hidden in the big data that are hard to decipher. Physics-based or first-principle-based theories are robust but suffer from the “curse of dimensionality” as the number of variables and degrees of freedom increases. Recently, many data-driven approaches particularly machine learning have shown advantages in dealing with high-dimensional problems, but they are usually agnostic and prone to unphysical failure.
Our research interest lies in seeking the intelligent integration of physics-based theories and data-driven approaches using our intersectional expertise in applied mechanics, electrochemistry, system engineering, and data science.
Thrust I. At micro & nanoscales, we learn physics (electro-chemo-mechanics) from experimental data.
Thrust II. At mesoscales, we implement known physics into computational models to discover the unknowns based on data.
Thrust III. At macroscales, we develop data-driven methods for the design and prognostics of real-world physics.
How can we incorporate intelligence?
We make use of state-of-the-art artificial intelligence techniques to efficiently generate the three Thrusts. For details, please follow our research publications & presentations.
Stay tuned for more updates!