1 code implementation • 28 Dec 2023 • Malik Hassanaly, Peter J. Weddle, Ryan N. King, Subhayan De, Alireza Doostan, Corey R. Randall, Eric J. Dufek, Andrew M. Colclasure, Kandler Smith
The techniques used to develop a PINN surrogate of the SPM are extended in Part II for the PINN surrogate for the P2D battery model, and explore the Bayesian calibration capabilities of both surrogates.
1 code implementation • 28 Dec 2023 • Malik Hassanaly, Peter J. Weddle, Ryan N. King, Subhayan De, Alireza Doostan, Corey R. Randall, Eric J. Dufek, Andrew M. Colclasure, Kandler Smith
To reduce the computational cost of Bayesian calibration, numerical solvers for physics-based models can be replaced with faster surrogates.
no code implementations • 22 Jun 2023 • Riccardo Balin, Filippo Simini, Cooper Simpson, Andrew Shao, Alessandro Rigazzi, Matthew Ellis, Stephen Becker, Alireza Doostan, John A. Evans, Kenneth E. Jansen
Recent years have seen many successful applications of machine learning (ML) to facilitate fluid dynamic computations.
1 code implementation • 25 May 2023 • Nuojin Cheng, Osman Asif Malik, Subhayan De, Stephen Becker, Alireza Doostan
An effective algorithm is proposed to maximize the variational lower bound of the HF log-likelihood in the presence of limited HF data, resulting in the synthesis of HF realizations with a reduced computational cost.
2 code implementations • 9 Nov 2022 • Kevin Doherty, Cooper Simpson, Stephen Becker, Alireza Doostan
We present a new convolution layer for deep learning architectures which we call QuadConv -- an approximation to continuous convolution via quadrature.
no code implementations • 3 Apr 2022 • Subhayan De, Matthew Reynolds, Malik Hassanaly, Ryan N. King, Alireza Doostan
Recent advances in modeling large-scale complex physical systems have shifted research focuses towards data-driven techniques.
no code implementations • 2 Feb 2022 • Aaron Allred, Lauren J. Abbott, Alireza Doostan, Kurt Maute
A new, machine learning-based approach for automatically generating 3D digital geometries of woven composite textiles is proposed to overcome the limitations of existing analytical descriptions and segmentation methods.
no code implementations • 31 Jan 2022 • Jacqueline Wentz, Alireza Doostan
Using results from PDE theory on coefficient decay rates, we construct an explicit generative model that predicts the polynomial chaos coefficient magnitudes.
no code implementations • 29 Jan 2022 • Alexandre Cortiella, Kwang-Chun Park, Alireza Doostan
In this work, we investigate and compare the performance of several local and global smoothing techniques to a priori denoise the state measurements and numerically estimate the state time-derivatives to improve the accuracy and robustness of two sparse regression methods to recover governing equations: Sequentially Thresholded Least Squares (STLS) and Weighted Basis Pursuit Denoising (WBPDN) algorithms.
no code implementations • 27 May 2021 • Subhayan De, Alireza Doostan
These bi-fidelity strategies are generalizations of transfer learning of neural networks that uses the parameters learned from a large low-fidelity dataset to efficiently train networks for a small high-fidelity dataset.
no code implementations • 30 Mar 2021 • Subhayan De, Bhuiyan Shameem Mahmood Ebna Hai, Alireza Doostan, Markus Bause
The physics model used in this study comprises of a monolithically coupled system of acoustic and elastic wave equations, known as the wave propagation in fluid-solid and their interface (WpFSI) problem.
no code implementations • 27 May 2020 • Alexandre Cortiella, Kwang-Chun Park, Alireza Doostan
The aim of this work is to improve the accuracy and robustness of SINDy in the presence of state measurement noise.
no code implementations • 11 Feb 2020 • Subhayan De, Jolene Britton, Matthew Reynolds, Ryan Skinner, Kenneth Jansen, Alireza Doostan
In the former approach, a neural network model mapping the inputs to the outputs of interest is trained based on the low-fidelity data.
1 code implementation • 11 Feb 2019 • Subhayan De, Jerrad Hampton, Kurt Maute, Alireza Doostan
To tackle this difficulty, we here propose an optimization approach that generates a stochastic approximation of the objective, constraints, and their gradients via a small number of adjoint (and/or forward) solves, per optimization iteration.
Optimization and Control Numerical Analysis
1 code implementation • 27 Jun 2016 • Paul G. Constantine, Alireza Doostan
Renewable energy researchers use computer simulation to aid the design of lithium ion storage devices.
Computational Physics Numerical Analysis