Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computational complexity”

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.

At least 361 records · Page 20
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Sylvester-preconditioned adaptive-rank implicit time integrators for advection-diffusion equations with variable coefficients

Here, we consider the adaptive-rank integration of multi-dimensional time-dependent advection-diffusion partial differential equations (PDEs) with variable coefficients. We employ a standard finite-difference method for spatial discretization coupled with high-order diagonally implicit Runge-Kutta temporal schemes. The discrete equation is a generalized Sylvester equation (GSE), which we solve with a projection-based adaptive-rank algorithm structured around two key strategies: (i) constructing dimension-wise subspaces using a novel atypical extended Krylov strategy, and (ii) efficiently solving the basis coefficient matrix with a preconditioned GMRES solver. The low-rank decomposition is performed in 2D using SVD and with high-order SVD (HOSVD) in 3D to represent the tensor in a compressed Tucker format. For d-dimensional problems (here, d = 2 or 3), the computational complexity and memory storage of the approach are found numerically to scale as and $\mathscr{O}(Nr^2) + \mathscr{O} (r^{d+1})$ and $\mathscr{O}(Nr) + \mathscr{O} (r^{d})$, respectively, with the one-dimensional resolution and the maximal rank during the Krylov iteration (which we find to be largely independent of on our numerical examples). We present numerical examples that illustrate the advertised properties of the algorithm.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Sub-grid drag model for immersed vertical cylinders in fluidized beds

Immersed vertical cylinders are often used as heat exchanger in gas-solid fluidized beds. Computational Fluid Dynamics (CFD) simulations are computationally expensive for large scale systems with bundles of cylinders. Therefore sub-grid models are required to facilitate simulations on a coarse grid, where internal cylinders are treated as a porous medium. The influence of cylinders on the gas-solid flow tends to enhance segregation and affect the gas-solid drag. A correction to gas-solid drag must be modeled using a suitable sub-grid constitutive relationship. In the past, Sarkar et al. have developed a sub-grid drag model for horizontal cylinder arrays based on 2D simulations. However, the effect of a vertical cylinder arrangement was not considered due to computational complexities. In this study, highly resolved 3D simulations with vertical cylinders were performed in small periodic domains. These simulations were filtered to construct a sub-grid drag model which can then be implemented in coarse-grid simulations. Gas-solid drag was filtered for different solids fractions and a significant reduction in drag was identified when compared with simulation without cylinders and simulation with horizontal cylinders. Slip velocities significantly increase when vertical cylinders are present. Lastly, vertical suspension drag due to vertical cylinders is insignificant however substantial horizontal suspension drag is observed which is consistent to the finding for horizontal cylinders.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Order parameter re-mapping algorithm for 3D phase field model of grain growth using FEM

Phase field modeling (PFM) is a well-known technique for simulating microstructural evolution. To model grain growth using PFM, typically each grain is assigned a unique non-conserved order parameter and each order parameter field is evolved in time. Traditional approaches using a one-to-one mapping of grains to order parameters present a challenge when modeling large numbers of grains due to the computational expense of using many order parameters. This problem is exacerbated when using an implicit finite element method (FEM), as the global matrix size is proportional to the number of order parameters. While previous work has developed methods to reduce the number of required variables and thus computational complexity and run time, none of the existing approaches can be applied for an implicit FEM implementation of PFM. Here, we present a modular, dynamic, scalable reassignment algorithm suitable for use in such a system. Polycrystal modeling with grain growth and stress require careful tracking of each grain’s position and orientation which is lost when using a reduced order parameter set. In conclusion, the method presented in this paper maintains a unique ID for each grain even after reassignment, to allow the PFM to be tightly coupled to calculations of the stress throughout the polycrystal. Implementation details and comparative results of our approach are presented.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Rapid sampling of stochastic displacements in Brownian dynamics simulations

Here, we present a new method for sampling stochastic displacements in Brownian Dynamics (BD) simulations of colloidal scale particles. The method relies on a new formulation for Ewald summation of the Rotne-Prager-Yamakawa (RPY) tensor, which guarantees that the real-space and wave-space contributions to the tensor are independently symmetric and positive-definite for all possible particle configurations. Brownian displacements are drawn from a superposition of two independent samples: a wave-space (far-field or long-ranged) contribution, computed using techniques from fluctuating hydrodynamics and non-uniform fast Fourier transforms; and a real-space (near-field or short-ranged) correction, computed using a Krylov subspace method. The combined computational complexity of drawing these two independent samples scales linearly with the number of particles. The proposed method circumvents the super-linear scaling exhibited by all known iterative sampling methods applied directly to the RPY tensor that results from the power law growth of the condition number of tensor with the number of particles. For geometrically dense microstructures (fractal dimension equal three), the performance is independent of volume fraction, while for tenuous microstructures (fractal dimension less than three), such as gels and polymer solutions, the performance improves with decreasing volume fraction. This is in stark contrast with other related linear-scaling methods such as the force coupling method and the fluctuating immersed boundary method, for which performance degrades with decreasing volume fraction. Calculations for hard sphere dispersions and colloidal gels are illustrated and used to explore the role of microstructure on performance of the algorithm. In practice, the logarithmic part of the predicted scaling is not observed and the algorithm scales linearly for up to 4×10 6 particles, obtaining speed ups of over an order of magnitude over existing iterative methods, and making the cost of computing Brownian displacements comparable to the cost of computing deterministic displacements in BD simulations. Finally, a high-performance implementation employing non-uniform fast Fourier transforms implemented on graphics processing units and integrated with the software package HOOMD-blue is used for benchmarking.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Reverse-mode differentiation in arbitrary tensor network format: with application to supervised learning.

This paper describes an efficient reverse-mode differentiation algorithm for contraction operations of tensor networks that may have arbitrary and unconventional network topologies. The approach leverages the tensor contraction tree of Evenbly and Pfeifer (2014), which provides an instruction set for the contraction sequence of a network. We show that this tree can be efficiently leveraged for differentiation of a full tensor network contraction using a recursive scheme that exploits (1) the bilinear property of contraction and (2) the property that trees have single path from root to leaves. While differentiation of tensor-tensor contraction is already possible in most automatic differentiation packages, we show that exploiting these two additional properties in the specific context of contraction sequences can improve efficiency. Following a description of the algorithm and computational complexity analysis, we investigate its utility for gradient-based supervised learning for low-rank function recovery and for fitting real-world unstructured datasets. We demonstrate improved performance over alternating least-squares optimization approaches and the capability to handle heterogeneous and arbitrary tensor network formats. When compared to alternating minimization algorithms, we find that the gradient-based approach requires a smaller oversampling ratio (number of samples compared to number model parameters) for recovery. This increased efficiency extends to fitting unstructured data of varying dimensionality and when employing a variety of tensor network formats. Here, we show improved learning using the hierarchical Tucker method over the tensor-train in high-dimensional settings on a number of benchmark problems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2013

RECOVERY ACT - Robust Optimization for Connectivity and Flows in Dynamic Complex Networks

The goal of this project was to study robust connectivity and flow patterns of complex multi-scale systems modeled as networks. Networks provide effective ways to study global, system level properties, as well as local, multi-scale interactions at a component level. Numerous applications from power systems, telecommunication, transportation, biology, social science, and other areas have benefited from novel network-based models and their analysis. Modeling and optimization techniques that employ appropriate measures of risk for identifying robust clusters and resilient network designs in networks subject to uncertain failures were investigated in this collaborative multi-university project. In many practical situations one has to deal with uncertainties associated with possible failures of network components, thereby affecting the overall efficiency and performance of the system (e.g., every node/connection has a probability of partial or complete failure). Some extreme examples include power grid component failures, airline hub failures due to weather, or freeway closures due to emergencies. These are also situations in which people, materials, or other resources need to be managed efficiently. Important practical examples include rerouting flow through power grids, adjusting flight plans, and identifying routes for emergency services and supplies, in the event network elements fail unexpectedly. Solutions that are robust under uncertainty, in addition to being economically efficient, are needed. This project has led to the development of novel models and methodologies that can tackle the optimization problems arising in such situations. A number of new concepts, which have not been previously applied in this setting, were investigated in the framework of the project. The results can potentially help decision-makers to better control and identify robust or risk-averse decisions in such situations. Formulations and optimal solutions of the considered problems need to capture uncertainty and risk using appropriate probabilistic, statistical and optimization concepts. The main difficulty arising in addressing these issues is the dramatic increase in the computational complexity of the resulting optimization problems. This project studied novel models and methodologies for risk-averse network optimization- specifically, network design, network flows and cluster detection problems under uncertainty. The approach taken was to incorporate a quantitative risk measure known as conditional value-at-risk that is widely used in financial applications. This approach presents a viable alternate modeling and optimization framework to chance-constrained optimization and mean-variance optimization, one that also facilitates the detection of risk-averse solutions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Applicability of available Li-ion battery degradation models for system and control algorithm design

Within electrified vehicle powertrains, lithium-ion battery performance degrades with aging and usage, resulting in a loss in both energy and power capacity. As a result, models used for system design and control algorithm development would ideally capture the impact of those efforts on battery capacity degradation, be computationally efficient, and simple enough to be used for algorithm development. This paper provides an assessment of the state-of-the-art in lithium-ion battery degradation models, including accuracy, computational complexity, and amenability to control algorithm development. Various aging and degradation models have been studied in the literature, including physics-based electrochemical models, semi-empirical models, and empirical models. Some of these models have been validated with experimental data; however, comparisons of pre-existing degradation models across multiple experimental data sets have not been previously published. Three representative models, a 1-d electrochemical model (a combination of performance model and degradation model), a semi-empirical degradation model (the performance is predicted by an equivalent circuit model) and an empirical degradation model (the performance is predicted by an equivalent circuit model), are compared against four published experimental data sets for a 2.3-Ah commercial graphite/LiFePO 4 cell. Based on simulation results and comparisons to experimental data, the key differences in the aging factors captured by each of the models are summarized. Here the results show that the physics-based model is best able to capture results across all four representative data sets with an error less than 10%, but is 20x slower than the empirical model, and 134x slower than the semi-empirical model, making it unsuitable for powertrain system design and model-based algorithm development. Despite being computationally efficient, the semi-empirical and empirical models, when used under conditions that lie outside the calibration data set, exhibit up to 71% error in capacity loss prediction. Such models require expensive experimental data collection to recalibrate for every new application. Thus, in the author’s opinion, there exists a need for a physically-based model that generalizes well across operating conditions, is computationally efficient for model-based design, and simple enough for control algorithm development.

25 ENERGY STORAGE↗
DOE OSTI2021

A Technical and Economic Assessment of LWR Flexible Operation for Generation and Demand Balancing to Optimize Plant Revenue

With increased penetration of subsidized variable renewable energy (VRE) resources and competition from low natural gas prices, existing light water reactor (LWR) nuclear power plants (NPPs) are struggling to remain economically competitive. This work examines the potential economic competitiveness of various thermal energy storage (TES) technologies when coupled directly or indirectly with a NPP. To highlight their relative economic competitiveness, we contrast several energy storage solutions in stochastic dispatch optimization. We leverage data from recent work analyzing a range of TES technologies with varying capital costs, performance, and technology readiness level (TRL) to establish our case. We explore inserting these technologies into an electricity market with existing nuclear generation and large projected variable renewable energy (VRE) penetration. Although these technologies' projected capital costs may make them unlikely candidates in their current state, this analysis demonstrates a high-fidelity techno-economic analysis of energy storage. Furthermore, as the projected cost of energy storage technologies evolves, this analysis sets a precedent for similar future investigations. One region with projected trends that may be unfavorable for existing nuclear capacity is the New York Independent System Operator (NYISO) market. New York state’s baseload generation has been historically provided by fossil-fired and nuclear assets. However, amid economic pressures from subsidized VREs and low natural gas prices, the state has recently deactivated Indian Point nuclear power plant units 2 and 3. Furthermore, the state plans to meet its zero-emission generation target by 2040 by replacing fossil-fired capacity with significant investments in VRE resources like wind and solar photovoltaic (PV) and battery storage. Increased intermittent resource penetration lowers the baseload power requirement, adding further economic pressure to the state’s three remaining NPPs still in operation. With three NPPs still in operation in New York, this work analyzes potential economic benefits to NPPs on the New York grid when directly or indirectly coupled with various TES technologies. This work requires two modeling steps to analyze the potential economic benefits of various system configurations of the TES directly or indirectly coupled with nuclear. First, this analysis leverages capacity expansion modeling by experts at the Electric Power Research Institute (EPRI). Using their deterministic capacity expansion model, U.S. Regional Economy, Greenhouse Gas, and Energy (US-REGEN), EPRI analysts evaluated the capacity and generation evolution of the New York state energy market under four projection scenarios. These four projection scenarios were developed to represent the potential evolution of the capacity and generation in NYISO from 2015 to 2050 under various economic, technology, and policy constraints. The results from these capacity expansion models are then used as boundary conditions in the second modeling step. The second modeling step uses the Holistic Energy Resource Optimization Network (HERON) for a set of stochastic techno-economic analyses (STEAs) to investigate the potential increase in the economic viability of various configurations of the TES. With no current capacity expansion capabilities, HERON takes the data generated from US-REGEN for 2050 to generate synthetic load, solar, and wind data. Then HERON economically optimizes the capacity and dispatch of the various TES configurations. The potential economic benefit is the differential net present value (NPV) of the TES configurations from the no-TES baseline. As a stochastic techno-economic analysis package, HERON introduces uncertainty into the economic metrics, while US-REGEN trades resolution for reduced computational complexity. Using HERON also allows the modeling of direct thermal coupling, a feature not common in capacity and dispatch models. As expected, with high capital costs, the costs of introducing energy storage for all the technologies considered outweighed the potential economic benefit of this strategy for flexible plant operation. The benefit of this analysis is primarily in demonstrating a workflow that examines innovative solutions to increase NPP revenue via TES coupling. HERON’s stochastic capacity and dispatch optimization process used in this work has proven an effective tool in observing and evaluating the impact of introducing storage technologies in a grid energy system.

25 ENERGY STORAGE↗
DOE OSTI2022

Efficient reconstruction and validation of heterogeneous microstructures for energy applications

The digital reconstruction of microstructures is necessary for simulations in fields ranging from geology to electrochemistry, but the state-of-the-art digital reconstruction techniques often compromise between resolution and field of view. It is challenging to retain detailed microstructure information in large-scale reconstructions. Here, this study investigates different aspects of the Yeong-Torquato algorithm based on correlation functions to make it more efficient. We achieve this goal by reducing the computational complexity of the chord-length distribution function and the two-point correlation function, applying the random sphere-packing method as the initial condition, and restricting potential voxel swaps to interfaces. In addition, a novel superposition parallel scheme is introduced to aid in searching for potential voxel swaps. The algorithm proposed is validated by comparing the pore-size distributions of reconstructed 3D custom battery electrodes from a sample dataset obtained from transmission X-ray microscopy. From a sample image with 200 x 200 pixels, the code can reconstruct a 300 x 300 x 300 structure in under 22 h and reconstruct a 400 x 400 x 400 structure in 43 h with eight cores.

42 ENGINEERING↗
DOE OSTI2017

ELSI: A unified software interface for Kohn–Sham electronic structure solvers

Solving the electronic structure from a generalized or standard eigenproblem is often the bottleneck in large scale calculations based on Kohn-Sham density-functional theory. This problem must be addressed by essentially all current electronic structure codes, based on similar matrix expressions, and by high-performance computation. We here present a unified software interface, ELSI, to access different strategies that address the Kohn-Sham eigenvalue problem. Currently supported algorithms include the dense generalized eigensolver library ELPA, the orbital minimization method implemented in libOMM, and the pole expansion and selected inversion (PEXSI) approach with lower computational complexity for semilocal density functionals. The ELSI interface aims to simplify the implementation and optimal use of the different strategies, by offering (a) a unified software framework designed for the electronic structure solvers in Kohn-Sham density-functional theory; (b) reasonable default parameters for a chosen solver; (c) automatic conversion between input and internal working matrix formats, and in the future (d) recommendation of the optimal solver depending on the specific problem. As a result, comparative benchmarks are shown for system sizes up to 11,520 atoms (172,800 basis functions) on distributed memory supercomputing architectures.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2017

The Approximability of Partial Vertex Covers in Trees.

Motivated by applications in risk management of computational systems, we focus our attention on a special case of the partial vertex cover problem, where the underlying graph is assumed to be a tree. Here, we consider four possible versions of this setting, depending on whether vertices and edges are weighted or not. Two of these versions, where edges are assumed to be unweighted, are known to be polynomial-time solvable (Gandhi, Khuller, and Srinivasan, 2004). However, the computational complexity of this problem with weighted edges, and possibly with weighted vertices, has not been determined yet. The main contribution of this paper is to resolve these questions, by fully characterizing which variants of partial vertex cover remain intractable in trees, and which can be efficiently solved. In particular, we propose a pseudo-polynomial DP-based algorithm for the most general case of having weights on both edges and vertices, which is proven to be NPhard. This algorithm provides a polynomial-time solution method when weights are limited to edges, and combined with additional scaling ideas, leads to an FPTAS for the general case. A secondary contribution of this work is to propose a novel way of using centroid decompositions in trees, which could be useful in other settings as well.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Improved treatment of exact exchange in Quantum ESPRESSO

Here, we present an algorithm and implementation for the parallel computation of exact exchange in Quantum ESPRESSO (QE) that exhibits greatly improved strong scaling. QE is an open-source software package for electronic structure calculations using plane wave density functional theory, and supports the use of local, semi-local, and hybrid DFT functionals. Wider application of hybrid functionals is desirable for the improved simulation of electronic band energy alignments and thermodynamic properties, but the computational complexity of evaluating the exact exchange potential limits the practical application of hybrid functionals to large systems and requires efficient implementations. We demonstrate that existing implementations of hybrid DFT that utilize a single data structure for both the local and exact exchange regions of the code are significantly limited in the degree of parallelization achievable. We present a band-pair parallelization approach, in which the calculation of exact exchange is parallelized and evaluated independently from the parallelization of the remainder of the calculation, with the wavefunction data being efficiently transformed on-the-fly into a form that is optimal for each part of the calculation. For a 64 water molecule supercell, our new algorithm reduces the overall time to solution by nearly an order of magnitude.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

On Bipartite Graphs Trees and Their Partial Vertex Covers.

Graphs can be used to model risk management in various systems. Particularly, Caskurlu et al. in [7] have considered a system, which has threats, vulnerabilities and assets, and which essentially represents a tripartite graph. The goal in this model is to reduce the risk in the system below a predefined risk threshold level. One can either restricting the permissions of the users, or encapsulating the system assets. The pointed out two strategies correspond to deleting minimum number of elements corresponding to vulnerabilities and assets, such that the flow between threats and assets is reduced below the predefined threshold level. It can be shown that the main goal in this risk management system can be formulated as a Partial Vertex Cover problem on bipartite graphs. It is well-known that the Vertex Cover problem is in P on bipartite graphs, however; the computational complexity of the Partial Vertex Cover problem on bipartite graphs has remained open. In this paper, we establish that the Partial Vertex Cover problem is NP-hard on bipartite graphs, which was also recently independently demonstrated [N. Apollonio and B. Simeone, Discrete Appl. Math., 165 (2014), pp. 37–48; G. Joret and A. Vetta, preprint, arXiv:1211.4853v1 [cs.DS], 2012]. We then identify interesting special cases of bipartite graphs, for which the Partial Vertex Cover problem, the closely related Budgeted Maximum Coverage problem, and their weighted extensions can be solved in polynomial time. We also present an 8/9-approximation algorithm for the Budgeted Maximum Coverage problem in the class of bipartite graphs. We show that this matches and resolves the integrality gap of the natural LP relaxation of the problem and improves upon a recent 4/5-approximation.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Final Report for Grant DOE DE-FG02-08ER64573 (Relating to Turbulence Modeling)

Turbulence enhances transport and mixing. Turbulent ocean currents can increase the encounter rate between fish larvae and their prey. Without turbulence, the fuel and air injected into the cylinder of an internal combustion engine would mix too slowly to be effective. Birds extract energy from turbulent winds to soar for great distances without flapping their wings. Turbulence can be detrimental as well. The efficiencies of vehicles, pipelines, and industrial equipment are all hindered by turbulence. Turbulence can also cause structural fatigue, generate unwanted noise, and distort the propagation of electromagnetic signals. These observations highlight the importance of turbulence research. Our ability to predict and control turbulence and, thus, to intensify or suppress its effects as circumstances warrant is contingent on our understanding of the underlying mechanisms. Turbulence is also immensely interesting from a purely scientific perspective and is a great source of fundamentally important, challenging problems for physicists, engineers, and mathematicians. Moreover, various methods and tools developed in the field have found applications in other fields, including nonlinear optics, nonlinear acoustics, pattern formation, image processing, data compression, and econophysics. It is possible, at least away from walls, to derive LES models of very high-order accuracy. However, these models also lead to correspondingly intense computational complexity for their numerical solution. Some also introduce yet more questions on the already difficult problem of specifying needed and sometimes artificial local boundary conditions for the inherently nonlocal flow averages. For these two (and other) reasons, there has been a resurgence of interest in basing simulations of turbulent flows on much simpler regularizations of the Navier–Stokes equations rather than on full models of local averages--i.e., LES models. Initially these were developed as pure theoretical tools. The influx of ideas from LES and the added constraint that the regularization be amenable to numerical simulation have breathed new perspective and thus new life into this thread of ideas.Over the past year, a significant portion of our research activity has focused on numerical studies of the Navier–Stokes-$αβ$ model and extensions thereof. Our results regarding these and other approaches to turbulence modeling are described below.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Predicting nonequilibrium Green’s function dynamics and photoemission spectra via nonlinear integral operator learning

Understanding the dynamics of nonequilibrium quantum many-body systems is an important research topic in a wide range of fields across condensed matter physics, quantum optics, and high-energy physics. However, numerical studies of large-scale nonequilibrium phenomena in realistic materials face serious challenges due to intrinsic high-dimensionality of quantum many-body problems and the absence of time-invariance. The nonequilibrium properties of many-body systems can be described by the dynamics of the correlator, or the Green's function of the system, whose time evolution is given by a high-dimensional system of integro-differential equations, known as the Kadanoff–Baym equations (KBEs). The time-convolution term in KBEs, which needs to be recalculated at each time step, makes it difficult to perform long-time numerical simulation. In this paper, we develop an operator-learning framework based on recurrent neural networks (RNNs) to address this challenge. We utilize RNNs to learn the nonlinear mapping between Green's functions and convolution integrals in KBEs. By using the learned operators as a surrogate model in the KBE solver, we obtain a general machine-learning scheme for predicting the dynamics of nonequilibrium Green's functions. Besides significant savings per each time step, the new methodology reduces the temporal computational complexity from $O(N_t^3)$ to $O(N_t)$ where N t is the number of steps taken in a simulation, thereby making it possible to study large many-body problems which are currently infeasible with conventional KBE solvers. Through various numerical examples, we demonstrate the effectiveness of the operator-learning based approach in providing accurate predictions of physical observables such as the reduced density matrix and time-resolved photoemission spectra. Moreover, our framework exhibits clear numerical convergence and can be easily parallelized, thereby facilitating many possible further developments and applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

High-order matrix-free incompressible flow solvers with GPU acceleration and low-order refined preconditioners

In this work, we present a matrix-free flow solver for high-order finite element discretizations of the incompressible Navier-Stokes and Stokes equations with GPU acceleration. For high polynomial degrees, assembling the matrix for the linear systems resulting from the finite element discretization can be prohibitively expensive, both in terms of computational complexity and memory. For this reason, it is necessary to develop matrix-free operators and preconditioners, which can be used to efficiently solve these linear systems without access to the matrix entries themselves. The matrix-free operator evaluations utilize GPU-accelerated sum-factorization techniques to minimize memory movement and maximize throughput. The preconditioners developed in this work are based on a low-order refined methodology with parallel subspace corrections, as described for diffusion problems in [1]. The saddle-point Stokes system is solved using block-preconditioning techniques, which are robust in mesh size, polynomial degree, time step, and viscosity. For the incompressible Navier-Stokes equations, we make use of projection (fractional step) methods, which require Helmholtz and Poisson solves at each time step. The performance of our flow solvers is assessed on several benchmark problems in two and three spatial dimensions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Bayesian Model Calibration for Extrapolative Prediction via Gibbs Posteriors

The current standard Bayesian approach to model calibration, which assigns a Gaussian process prior to the discrepancy term, often suffers from issues of unidentifiability and computational complexity and instability. When the goal is to quantify uncertainty in physical parameters for extrapolative prediction, then there is no need to perform inference on the discrepancy term. With this in mind, we introduce Gibbs posteriors as an alternative Bayesian method for model calibration, which updates the prior with a loss function connecting the data to the parameter. The target of inference is the physical parameter value which minimizes the expected loss. We propose to tune the loss scale of the Gibbs posterior to maintain nominal frequentist coverage under assumptions of the form of model discrepancy, and present a bootstrap implementation for approximating coverage rates. Our approach is highly modular, allowing an analyst to easily encode a wide variety of such assumptions. Furthermore, we provide a principled method of combining posteriors calculated from data subsets. We apply our methods to data from an experiment measuring the material properties of tantalum.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Score-Based Physics-Informed Neural Networks for High-Dimensional Fokker–Planck Equations

The Fokker-Planck (FP) equation is a foundational partial differential equation (PDE) in stochastic processes involving Brownian motions. However, the curse of dimensionality (CoD) poses a formidable challenge when dealing with high-dimensional FP equations. Although Monte Carlo simulation and (vanilla) Physics-Informed Neural Networks (PINNs) have shown the potential to tackle CoD, both methods exhibit significant numerical errors in high dimensions when dealing with the probability density function (PDF) associated with Brownian motion. The point-wise PDF values tend to decrease exponentially as dimensionality increases, surpassing the precision of numerical simulations and resulting in substantial errors. In addition, due to its massive sampling, Monte Carlo fails to offer fast sampling. Modeling the logarithm likelihood (LL) via vanilla PINNs transforms the FP equation into a notoriously difficult Hamilton-Jacobi-Bellman (HJB) equation, which is impractical for PINN learning, whose error grows rapidly with dimension. To this end, we propose a novel approach utilizing a score-based solver to fit the score function in stochastic differential equations (SDEs). The score function, defined as the gradient of the LL, plays a fundamental role in inferring LL and PDF and enables fast SDE sampling, offering an effective means to overcome the CoD. Three fitting methods, Score Matching (SM), Sliced Score Matching (SSM), and Score-PINN, are introduced, each contributing unique advantages in computational complexity, accuracy, and generality. The proposed score-based SDE solver operates in two stages: first, employing score matching or Score-PINN to acquire the score function; and second, solving the LL via an ordinary differential equation (ODE) using the obtained score function. Comparative evaluations across these methods showcase varying trade-offs. The proposed methodology is evaluated across diverse SDEs, including anisotropic Ornstein-Uhlenbeck processes, geometric Brownian motion, and Brownian motion with varying eigenspace. We also test various distributions, including Gaussian, Log-normal, Laplace, and Cauchy distributions. The numerical results demonstrate the score-based SDE solver’s stability, speed, and performance across different experimental settings, solidifying its potential as a solution to CoD for high-dimensional FP equations.

97 MATHEMATICS AND COMPUTING↗