Optimal Experiment Design for Large-Scale Inverse Problems with Enhanced Robustness to Model Uncertainties
Explore the source record for details and available documents.
Smith, Chandler Baldwin [Sandia National Laborator↗
SEARCH · Search NASA
Results for “optimization, inverse problems”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Algorithm Solves Constrained and Unconstrained Optimization Problems
Is quasi-Newton iteration utilizing Broyden/Fletcher/Goldfarb/Shanno update on inverse Hessian matrix. Capable of solving constrained optimization unconstrained optimization and constraints only problems with one to five independent variables from one to five constraint functions and one dependent function optimized.
Denson, M. A.↗
LaSDI: Parametric Latent Space Dynamics Identification
Enabling fast and accurate physical simulations with data has become an important area of computational physics to aid in inverse problems, design-optimization, uncertainty quantification, and other various decision-making applications. This paper presents a data-driven framework for parametric latent space dynamics identification procedure that enables fast and accurate simulations. The parametric model is achieved by building a set of local latent space model and designing an interaction among them. An individual local latent space dynamics model achieves accurate solution in a trust region. By letting the set of trust region to cover the whole parameter space, our model shows an increase in accuracy with an increase in training data. Herein we introduce two different types of interaction mechanisms, i.e., point-wise and region-based approach. Both linear and nonlinear data compression techniques are used. We illustrate the framework of Latent Space Dynamics Identification (LaSDI) enable a fast and accurate solution process on various partial differential equations, i.e., Burgers’ equations, radial advection problem, and nonlinear heat conduction problem, achieving 0 (100)x speed-up and 0 (1)% relative error with respect to the corresponding full order models.
97 MATHEMATICS AND COMPUTING↗
Evaluation of dual-weighted residual and machine learning error estimation for projection-based reduced-order models of steady partial differential equations
Projection-based reduced-order models (pROMs) show great promise as a means to accelerate many-query applications such as forward error propagation, solving inverse problems, and design optimization. In order to deploy pROMs in the context of high-consequence decision making, accurate error estimates are required to determine the region(s) of applicability in the parameter space. The following paper considers the dual-weighted residual (DWR) error estimate for pROMs and compares it to another promising pROM error estimate, machine learned error models (MLEM). Here, we show how DWR can be applied to ROMs and then evaluate DWR on two partial differential equations (PDEs): a two-dimensional linear convection–reaction–diffusion equation, and a three-dimensional static hyper-elastic beam. It is shown that DWR is able to estimate errors for pROMs extrapolating outside of their training set while MLEM is best suited for pROMs used to interpolate within the pROM training set.
42 ENGINEERING↗
TRINIDI: Time-of-Flight Resonance Imaging With Neutrons for Isotopic Density Inference
Accurate reconstruction of 2D and 3D isotope densities is a desired capability with great potential impact in applications such as evaluation and development of next-generation nuclear fuels. Neutron time-of-flight (TOF) resonance imaging offers a potential approach by exploiting the characteristic neutron absorption spectra of each isotope. However, it is a major challenge to compute quantitatively accurate images due to a variety of confounding effects such as severe Poisson noise, background scatter, beam non-uniformity, absorption non-linearity, and extended source pulse duration. We present the TRINIDI algorithm which is based on a two-step process in which we first estimate the neutron flux and background counts, and then reconstruct the areal densities of each isotope and pixel. Both components are based on the inversion of a forward model that accounts for the highly non-linear absorption, energy-dependent emission profile, and Poisson noise, while also modeling the substantial spatio-temporal variation of the background and flux. Further, to do this, we formulate the non-linear inverse problem as two optimization problems that are solved in sequence. We demonstrate on both synthetic and measured data that TRINIDI can reconstruct quantitatively accurate 2D views of isotopic areal density that can then be reconstructed into quantitatively accurate 3D volumes of isotopic volumetric density.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
High-Dimensional Spectral Sampling
A wide variety of applications in science and engineering depend on effectively and efficiently exploring high- dimensional parameter spaces. Common examples are uncertainty quantification, inverse problems, or various optimization approaches. In such applications, the first step is to create an initial uniform, random sampling of the space to create a baseline of knowledge. The simulations evaluated at these samples are then used for subsequent processing. In its most generic form, the goal of sampling is to produce the maximal amount of information with the minimal number of samples. However, despite being the first step and thus influencing the performance of the entire workflow, the quality of the initial sample designs is rarely analyzed or optimized. This project develops a new spectral sampling theory for analyzing and creating experiment designs in high-dimensional spaces. Further, new algorithms for estimating sample properties and sample synthesis are designed. Finally, applications of the proposed spectral sampling tools in surrogate modeling, scientific machine learning and image analysis are presented.
97 MATHEMATICS AND COMPUTING↗
Stepwise Dynamic Calibration of a Hydromechanical Simulation Using Time-Lapse Vertical Seismic Profile
This study aims to develop a methodology for calibrating subsurface stress changes through time-lapse Vertical Seismic Profiling (VSP) integration. The selected study site is the 13-10A injector well within the ongoing CO2-EOR operation of the Farnsworth Field Unit. The Time-lapse VSP dataset carries the combined effects of fluid substitution and mean effective stress changes, thereby providing a dataset amenable for the calibration of production and injection-induced stress changes. The concept is similar to calibrating a reservoir simulation model in that the process honor real field data to set up an inverse problem. The solution optimizes the independent and impactful geomechanical parameters that replicate the observed time-lapse seismic velocity changes. This stress calibration is enabled by 4D geomechanical modeling and the VSP Integration workflow. This calibration benefits from extensive geological, geophysical and geomechanical characterization through 3D seismic data, geophysical well logs, and core assessed as part of the 1D MEM conducted on the 13-10A subject well. These data are used to develop a site-specific rock physics model. The Biot Gassmann workflow combines rock physics and reservoir simulation outputs to determine the fluid substitution contribution to seismic velocity change. Additionally, modeled seismic velocity attributed to mean effective stress are determined from the geomechanical simulation outputs, and the stress-velocity relationship developed from the ultrasonic seismic velocity measurements on the extracted Morrow B core. A penalty function is then formed between the modeled seismic velocities and the observed time-lapse VSP dataset. Four independent and impactful geomechanical parameters have been determined. These are the bulk modulus and shear modulus for zero porosity and the shear and compressional seismic velocity to mean effective stress derivatives. The dataset of numerous coupled hydromechanical- geomechanical simulation realizations is built by combining variations of the four stated geomechanical parameters. A machine learning-assisted workflow comprised of an artificial neural network and a particle swarm optimizer are used to converge on the optimal geomechanical parameters. The successful execution of this workflow has affirmed the suitability of acoustic time-lapse measurements for 4D-VSP geomechanical stress calibration pending measurable stress sensitivities within the anticipated effective stress changes and the availability of suitable and reliable datasets for petroelastic modeling.
02 PETROLEUM↗
Bilevel optimization, deep learning and fractional Laplacian regularization with applications in tomography
Here we consider a generalized bilevel optimization framework for solving inverse problems. We introduce fractional Laplacian as a regularizer to improve the reconstruction quality, and compare it with the total variation regularization. We emphasize that the key advantage of using fractional Laplacian as a regularizer is that it leads to a linear operator, as opposed to the total variation regularization which results in a nonlinear degenerate operator. Inspired by residual neural networks, to learn the optimal strength of regularization and the exponent of fractional Laplacian, we develop a dedicated bilevel optimization neural network with a variable depth for a general regularized inverse problem. We illustrate how to incorporate various regularizer choices into our proposed network. As an example, we consider tomographic reconstruction as a model problem and show an improvement in reconstruction quality, especially for limited data, via fractional Laplacian regularization. We successfully learn the regularization strength and the fractional exponent via our proposed bilevel optimization neural network. We observe that the fractional Laplacian regularization outperforms total variation regularization. This is specially encouraging, and important, in the case of limited and noisy data.
97 MATHEMATICS AND COMPUTING↗
Where did the tumor start? An inverse solver with sparse localization for tumor growth models
In this work, we present a numerical scheme for solving an inverse problem for parameter estimation in tumor growth models for glioblastomas, a form of aggressive primary brain tumor. The growth model is a reaction–diffusion partial differential equation (PDE) for the tumor concentration. We use a PDE-constrained optimization formulation for the inverse problem. The unknown parameters are the reaction coefficient (proliferation), the diffusion coefficient (infiltration), and the initial condition field for the tumor PDE. Segmentation of magnetic resonance imaging (MRI) scans drive the inverse problem where segmented tumor regions serve as partial observations of the tumor concentration. Like most cases in clinical practice, we use data from a single time snapshot. Moreover, the precise time relative to the initiation of the tumor is unknown, which poses an additional difficulty for inversion. We perform a frozen-coefficient spectral analysis and show that the inverse problem is severely ill-posed. We introduce a biophysically motivated regularization on the structure and magnitude of the tumor initial condition. In particular, we assume that the tumor starts at a few locations (enforced with a sparsity constraint on the initial condition of the tumor) and that the initial condition magnitude in the maximum norm is equal to one. We solve the resulting optimization problem using an inexact quasi-Newton method combined with a compressive sampling algorithm for the sparsity constraint. Our implementation uses PETSc and AccFFT libraries. We conduct numerical experiments on synthetic and clinical images to highlight the improved performance of our solver over a previously existing solver that uses standard two-norm regularization for the calibration parameters. The existing solver is unable to localize the initial condition. Our new solver can localize the initial condition and recover infiltration and proliferation. In clinical datasets (for which the ground truth is unknown), our solver results in qualitatively different solutions compared to the two-norm regularized solver.
97 MATHEMATICS AND COMPUTING↗
Learning Optimal Aerodynamic Designs
This project created a framework for efficient, accurate, and scalable deep neural network representations of design optimization problem solutions. The inputs to these DNN representations are the vector of design requirement parameters, the outputs are the optimal design variables, and the goal is to learn the map from inputs to outputs (i.e., inverse design). The team addressed the problem of the optimal shape design of aerodynamic lifting surfaces—in particular aircraft wings—using a Reynolds-Average Navier Stokes model to govern the CFD-based aerodynamic shape optimization. The inverse design map for such problems is very complex and high-dimensional, involving inputs and outputs on the order of 1000s. To approximate this inverse design map, the team developed algorithms to construct parsimonious DNN architectures, which automatically identify low-dimensional manifolds in which design requirements affect optimal shape parameters, and trained these architectures with multifidelity optimization methods. The resulting methodology accurately and automatically designs optimal aerodynamic lifting surfaces with very high accuracy (99%) at interactive speeds, of the order of milliseconds, resulting in factors of one million or more speedup relative to CFD-based design optimization.
97 MATHEMATICS AND COMPUTING↗
Improve earthquake hypocenter using adaptive simulated annealing inversion in regional tectonic, volcano tectonic, and geothermal observation
Observation of earthquakes is routinely used widely in tectonic activity observation, and also in local scale such as volcano tectonic and geothermal activity observation. It is necessary for determining the location of precise hypocenter which the process involves finding a hypocenter location that has minimum error between the observed and the calculated travel times. When solving this nonlinear inverse problem, simulated annealing inversion method can be applied to such global optimization problems, which the convergence of its solution is independent of the initial model. In this study, we developed own program codeby applying adaptive simulated annealing inversion in Matlab environment. We applied this method to determine earthquake hypocenter using several data cases which are regional tectonic, volcano tectonic, and geothermal field. The travel times were calculated using ray tracing shooting method. We then compared its results with the results using Geiger’s method to analyze its reliability. Our results show hypocenter location has smaller RMS error compared to the Geiger’s result that can be statistically associated with better solution. The hypocenter of earthquakes also well correlated with geological structure in the study area. Werecommend using adaptive simulated annealing inversion to relocate hypocenter location in purpose to get precise and accurate earthquake location.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Drive-pressure optimization in ramp-wave compression experiments through differential evolution
Ramp-wave dynamic-compression experiments are used to examine quasi-isentropic loading paths in materials. The gradual and continuous increase in pressure created by ramp waves make these types of experiments ideal for studying nonequilibrium material behavior, such as solidification kinetics. In ramp-wave compression experiments, the input drive pressure to the experimental setup may be exerted through one of a number of different mechanisms (e.g., magnetic fields, gas-gun-driven impactors, or high-energy lasers) and is generally required for simulating such experiments. Yet, regardless of the specific mechanism, this drive pressure cannot be measured directly (measurements are generally taken at a location near the back of the experimental setup through a transparent window), leading to an inverse problem where one must determine the drive pressure at the front of the experimental setup (i.e., the input) that corresponds to the particle velocity (the output) measured near the back of the experimental setup. Furthermore, we solve this inverse problem using a heuristic optimization algorithm, known as differential evolution, coupled with a multiphysics, hydrodynamics code that simulates the compression of the experimental setup. By running many rounds of forward simulations of the experimental setup, our optimization process iteratively searches for a drive pressure that is optimized to closely reproduce the experimentally measured particle velocity near the back of the experimental setup. While our optimization methodology requires a significant number of hydrodynamics simulations to be conducted, many of these can be performed in parallel, which greatly reduces the time cost of our methodology. One novel aspect of our method for determining the drive pressure is that it does not require physical modeling of the drive mechanism and can thus be broadly applied to many types of ramp-compression experiments, regardless of the drive mechanism.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
Multigrid Optimization for Large-Scale Ptychographic Phase Retrieval
Ptychography is a popular imaging technique that combines diffractive imaging with scanning microscopy. The technique consists of a coherent beam that is scanned across an object in a series of overlapping positions, leading to reliable and improved reconstructions. Ptychographic microscopes allow for large fields to be imaged at high resolution at the cost of additional computational expense. Herein, we propose a multigrid-based optimization framework to reduce the computational burdens of large-scale ptychographic phase retrieval. Our proposed method exploits the inherent hierarchical structures in ptychography through tailored restriction and prolongation operators for the object and data domains. Our numerical results show that our proposed scheme accelerates the convergence of its underlying solver and outperforms the ptychographical iterative engine, a workhorse in the optics community.
97 MATHEMATICS AND COMPUTING↗
Grad–Shafranov equilibria via data-free physics informed neural networks
A large number of magnetohydrodynamic (MHD) equilibrium calculations are often required for uncertainty quantification, optimization, and real-time diagnostic information, making MHD equilibrium codes vital to the field of plasma physics. In this paper, we explore a method for solving the Grad–Shafranov equation by using physics-informed neural networks (PINNs). For PINNs, we optimize neural networks by directly minimizing the residual of the partial differential equation as a loss function. We show that PINNs can accurately and effectively solve the Grad–Shafranov equation with several different boundary conditions, making it more flexible than traditional solvers. This method is flexible as it does not require any mesh and basis choice, thereby streamlining the computational process. We also explore the parameter space by varying the size of the model, the learning rate, and boundary conditions to map various tradeoffs such as between reconstruction error and computational speed. Additionally, we introduce a parameterized PINN framework, expanding the input space to include variables such as pressure, aspect ratio, elongation, and triangularity in order to handle a broader range of plasma scenarios within a single network. Parameterized PINNs could be used in future work to solve inverse problems such as shape optimization.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Estimation of extreme temperatures in direct solar methane pyrolysis within a porous medium
Porous media have wide application in renewable energy conversion processes, such as solar-thermal fuels production and decarbonization. Heat transport mechanisms within porous media can be highly complex, particularly under extreme conditions encountered in concentrated solar thermal reactors in which direct measurement of temperature is challenging. Here, we implement and report an inverse heat conduction model to estimate the temperature distribution throughout a porous substrate domain in a direct solar methane pyrolysis process. By solving a two-dimensional heat transfer problem and applying an inverse optimization algorithm, we estimate the quasi-steady state spatial temperature distribution in a fibrous porous carbon substrate. The results are validated indirectly by experimentally measured graphite deposition and a simplified reaction kinetic model.
finite difference method↗
An introduction to partial differential equations constrained optimization
Partial differential equation (PDE) constrained optimization is designed to solve control, design, and inverse problems with underlying physics. A distinguishing challenge of this technique is the handling of large numbers of optimization variables in combination with the complexities of discretized PDEs. Over the last several decades, advances in algorithms, numerical simulation, software design, and computer architectures have allowed for the maturation of PDE constrained optimization (PDECO) technologies with subsequent solutions to complicated control, design, and inverse problems. This special journal edition, entitled “PDE-Constrained Optimization”, features eight papers that demonstrate new formulations, solution strategies, and innovative algorithms for a range of applications. In particular, these contributions demonstrate the impactfulness on our engineering and science communities. This paper offers short remarks to provide some perspective and background for PDECO, in addition to summaries of the eight papers.
97 MATHEMATICS AND COMPUTING↗
Solving inverse problems for process-structure linkages using asynchronous parallel Bayesian optimization.
Abstract not provided.
Tran, Anh↗
MatCal Users Guide: Release 1.3.0
Any continuum mechanics model will require three components: (1) a discretized geometry of the boundary value problem being studied, (2) the partial differential equations to be solved, and (3) the initial conditions and boundary conditions for the problem. To describe material behavior in these computational models, material models contribute to (2) the underlying equations and, occasionally, to (3) the initial conditions for the simulation. These material models can exhibit a mathematical form that is empirically based, based on first principles, or developed from both empirical observations and known physics. In general, these models are meant to represent a class of materials with well understood behavior. As a result, material models have parameters that must be tuned or calibrated so that the model response matches characterization data available for the specific material it is intended to represent when used to simulate a specific system. For simple models, such as isotropic, linear elastic materials in solid mechanics, this calibration process can be a simple analytical calculation directly extracting the parameters from experimental measurements. For complex models that have many inputs and require many characterization datasets to adequately identify the material behavior, the model calibration process can require an inverse problem approach where an optimization is performed to tune the model parameters to the available data.
36 MATERIALS SCIENCE↗