Scientific Computing and Matrix Computations Seminar: Nonlinear reduced-order modeling: Using machine learning to enable extreme-scale simulation for many-query problems
Seminar: Scientific Computing: CS | April 17 | 2-3 p.m. | 380 Soda Hall
Kevin Carlberg, Sandia Lab
Physics-based modeling and simulation has become indispensable across many applications in engineering and science, ranging from aircraft design to monitoring national critical infrastructure. However, as simulation is playing an increasingly important role in scientific discovery, decision making, and design, greater demands are being placed on model fidelity. This high fidelity necessitates modeling fine spatiotemporal resolution, which can lead to extreme-scale, nonlinear dynamical-system models whose simulations consume months on thousands of computing cores. Further, most scientific applications (e.g., uncertainty quantification, design optimization) are many query in nature, as they require the (parameterized) model to be simulated thousands of times. This leads to a computational barrier: the computational cost of high-fidelity simulations renders them impractical for many-query problems.
In this talk, I will present several advances in the field of nonlinear model reduction that exploit simulation data to overcome this barrier. These methods combine concepts from machine learning, computational physics, and numerical optimization to produce low-dimensional reduced-order models (ROMs) that are 1) accurate, 2) low cost, 3) structure preserving, 4) reliable, and 5) certified. First, I will describe least-squares PetrovGalerkin projection, which leverages subspace identification and optimal projection to ensure accuracy. Second, I will describe the sample mesh concept, which employs empirical regression and greedy-optimal sensor-placement techniques to ensure low cost. I will also describe novel methods that exploit time-domain data to further reduce computational costs. Third, I will present a technique that ensures the ROM is globally conservative in the case of finite-volume discretizations, thus ensuring structure preservation. Fourth, I will describe model reduction on nonlinear manifolds, wherein we employ convolutional autoencoders from deep learning to improve the predictive accuracy of the ROM. I will also describe ROM h-adaptivity, which employs concepts from goal-oriented mesh adaptivity to ensure that the ROM is reliable, i.e., it can satisfy any prescribed error tolerance. Finally, I will present machine-learning error models, which apply regression methods from machine learning to construct a statistical model for the ROM error; this quantifies the ROM-induced epistemic uncertainty and provides a mechanism for certification.