Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “algorithmic differentiation”

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 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

STILGAR End-of-Project Report

The Subsurface Tunnel Imaging LeveraGed by Analysis of Rayleigh wave ellipticity (STILGAR) project demonstrated an integrated geophysical approach for detecting, locating, and characterizing underground structural changes using dense seismic arrays and advanced inversion techniques. Field campaigns were conducted at two operational mines—the Redmond salt mine (Utah) and Graymont Pleasant Gap limestone mine (Pennsylvania)—providing real-world testbeds for monitoring anthropogenic subsurface activity. At the Redmond salt mine, seismic interferometry combined with back-projection inversion successfully identified continuous, low-amplitude signals from mining operations. The approach differentiated stationary from migrating anthropogenic sources, captured daily operational cycles, and validated the potential of passive seismic monitoring for remote detection of underground activity. At the Graymont Pleasant Gap mine, two dense seismic deployments in the spring and fall of 2023 generated over 4 TB of high-resolution data. Key outcomes included the relocation of 199 underground and 8 surface explosions with accuracies within tens of meters and the development of a 3D P-wave velocity model using the triple-difference tomography algorithm (tomoTD) that resolved major structural features such as the mine entrance, low-velocity tunnels, and roof-collapse areas. Ambient noise cross-correlation and back-projection analyses revealed persistent sources linked to ongoing mining activity, whereas horizontal-to-vertical spectral ratio (HVSR) and ellipticity studies confirmed stable site responses across seasons and identified soil thickness trends consistent with regional erosional and depositional processes. Checkerboard and sensitivity tests further validated the robustness of the tomographic results. Overall, the findings emphasize that although significant progress has been made in subsurface imaging, further work is needed to enhance the detection and localization of underground structures. Accurate imaging requires higher frequencies, yet anthropogenic sources tend to dominate the seismic record at those frequencies, and high-frequency surface waves are affected by higher modes that complicate interpretation. The improved detection and localization of human-induced signals enabled detailed temporal and spatial mapping of daily mine operations, demonstrating the feasibility of continuous anthropogenic source monitoring. Sensitivity to signals from nontraditional sources, such as fan operations, highlights the broader applicability of this approach to other industrial environments in which continuous and impulsive signals are present. The field campaigns produced a substantial volume of high-quality seismic data, supporting the development and testing of new methods for seismic source characterization and subsurface imaging. Future deployments should include sensors capable of recording lower frequencies to probe deeper structures, increase bandwidth to enhance resolution and sensitivity to both shallow and deep targets, and collect additional large-scale datasets to refine imaging and source characterization techniques. Moreover, conducting 3D modeling studies of seismic wavefields at higher frequencies will provide a better understanding of wave scattering and cavity–wavefield interactions in complex underground environments. In conclusion, the STILGAR project demonstrated that integrated seismic monitoring can effectively characterize underground operations, capturing both natural and anthropogenic signals. The approaches developed provide a foundation for improved detection, localization, and imaging of subsurface structures and are directly transferable to broader industrial monitoring applications.

58 GEOSCIENCES↗
DOE OSTI2020

Simultaneous material, shape and topology optimization

Using three design fields we develop an optimization environment that can simultaneously optimize material, shape and topology. We use the implicit representation of the boundaries with level-set functions that define the shape and topology. Differentiable R-functions allow us to combine these shapes and topology descriptions with Boolean operations. Additionally, we incorporate design dependent-stiffness materials with another design field. Notably, this framework accommodates design dependent loads, has the ability to introduce holes, and ensures the satisfaction of optimality criteria. It builds upon the fictitious domain, ersatz material, material interpolation and level-set methods. Additionally, it also borrows from parameterized density-based topology optimization methods. Since analytical sensitivities can be computed, we use efficient nonlinear programming algorithms to update the design instead of the Hamilton–Jacobi’s scheme of level-set methods. We illustrate the features of our framework by designing a cantilever beam with octet truss microlattice, a dam with design-dependent loads, and a composite clevis plate.

42 ENGINEERING↗
DOE OSTI2025

Kinetic Deep Learning v0.1

Here, we present a method that uses protein levels to predict times series of metabolite concentrations. Understanding this type of pathway dynamics is important in order to predict the behavior of the pathway and, more pragmatically, to be able to design biological systems (such as strains bioengineered to produce chemical products) reliably. Typically, for this purpose, kinetic models consisting of differential equations based on the Michaelis-Menten dynamics have been used in the past. However, these methods can rarely produce good fits to measured data time series. Possibly, this happens because the kinetic constants are unknown or are different from the ones measured in vivo, or perhaps because Michaelis-Menten dynamics is not a satisfactory description. In order to improve the predictive nature of these kinetic models we have eliminated the Michaelis-Menten description of pathway dynamics and we have substituted it by algorithms that automatically learn these dynamics from previously obtained metabolomics and proteomics data using machine learning approaches. Specifically, kinetic deep learning uses deep learning to map proteomics time series to metabolite concentration time series, instead of learning the first metabolite derivative and integrating in (as in the first version of kinetic learning). This approach is shown to provide good to excellent results with a data set specifically collected for this purpose.

Garcia Martin, Hector [Joint BioEnergy Institute (↗
DOE OSTI2023

Projecting Future Energy Production from Operating Wind Farms in North America. Part II: Statistical Downscaling

Abstract Capacity factors (CFs) derived from daily expected power at 22 operating wind farms in different regions of North America are used as predictands to train statistical downscaling algorithms using output from ERA5. The statistical downscaling models are then used to make CF projections for a suite of CMIP6 Earth System Models (ESMs). Downscaling is performed using a hybrid statistical approach that employs synoptic types derived using k -means clustering applied to sea level pressure fields with variance corrections applied as a function of the pressure gradient intensity. ESMs exhibit marked variability in terms of the skill with which the frequency of synoptic types and pressure gradients are reproduced relative to ERA5, and that differential skill is used to infer differential credibility in the associated CF projections. Projections of median annual mean CF [P50(CF)] in each 20-yr period from 1980 to 2099 show evidence of declines at most wind farms except in parts of the southern Great Plains, although the magnitude of the changes is strongly dependent on the ESM. For example, P50(CF) in 2080–99 deviate from those in 1980–99 by from −3.1 to +0.2 percentage points in the Northeast. The largest-magnitude declines in P50(CF) ranging from −3.9 to −2 percentage points are projected for the southern West Coast. CF trends exhibit marked seasonality and are strongly linked to changes in the relative intensity of future synoptic patterns, with much less impact from shifts in the occurrence of synoptic types over time. Internal climate modes continue to play a significant role in inducing interannual variability in wind power production, even under high radiative forcing scenarios. Significance Statement We describe how future climate changes may affect wind resources and wind power generation. Near-term changes in projected wind power electricity generation potential at operating wind farms over North America are small, but by the end of the current century electricity production is projected to decrease in many areas but may increase in parts of the southern Great Plains. The amount of change in projected wind power production is a strong function of the Earth system model that is downscaled and also depends on the continued presence of internally forced climate variability. An additional dependence on the amount of greenhouse gas–induced global warming indicates the transition of the energy sector to low-carbon sources may assist in maintaining the abundant U.S. wind resource.

Meteorology & Atmospheric Sciences↗
DOE OSTI2025

Generative AI models for learning flow maps of stochastic dynamical systems in bounded domains

Simulating stochastic differential equations (SDEs) in bounded domains, presents significant computational challenges due to particle exit phenomena, which requires accurate modeling of interior stochastic dynamics and boundary interactions. Despite the success of machine learning-based methods in learning SDEs, existing learning methods are not applicable to SDEs in bounded domains because they cannot accurately capture the particle exit dynamics. We present a unified hybrid data-driven approach that combines a conditional diffusion model with an exit prediction neural network to capture both interior stochastic dynamics and boundary exit phenomena. Our ML model consists of two major components: a neural network that learns exit probabilities using binary cross-entropy loss with rigorous convergence guarantees, and a training-free diffusion model that generates state transitions for non-exiting particles using closed-form score functions. The two components are integrated through a probabilistic sampling algorithm that determines particle exit at each time step and generates appropriate state transitions. Here, the performance of the proposed approach is demonstrated via three test cases: a one-dimensional simplified problem for theoretical verification, a two-dimensional advection-diffusion problem in a bounded domain, and a three-dimensional problem of interest to magnetically confined fusion plasmas.

Bounded domains↗
DOE OSTI2025

Computing Nonequilibrium Responses with Score-Shifted Stochastic Differential Equations

Using equilibrium fluctuations to understand the response of a physical system to an externally imposed perturbation is the basis for linear response theory, which is widely used to interpret experiments and shed light on microscopic dynamics. For nonequilibrium systems, perturbations cannot be interpreted simply by monitoring fluctuations in a conjugate observable and general response results rely on path ensemble averaging. Furthermore, these techniques do not apply to perturbations that affect the diffusion tensor in a stochastic system. Here, we introduce an “effective” physical process that represents the diffusion perturbed dynamics and enables accurate calculations of responses to a change in the diffusion. Interestingly, the effective dynamics contain an additional drift involving the instantaneous “score” of the system, and we leverage score matching algorithms to carry out nonequilibrium response calculations on systems for which the exact stationary distribution is unknown.

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

A Three-Dimensional Reconstruction Algorithm for Scanning Transmission Electron Microscopy Data from a Single Sample Orientation

Abstract Increasing interest in three-dimensional nanostructures adds impetus to electron microscopy techniques capable of imaging at or below the nanoscale in three dimensions. We present a reconstruction algorithm that takes as input a focal series of four-dimensional scanning transmission electron microscopy (4D-STEM) data. We apply the approach to a lead iridate, PbIrO, and yttrium-stabilized zirconia, YZrO, heterostructure from data acquired with the specimen in a single plan-view orientation, with the epitaxial layers stacked along the beam direction. We demonstrate that Pb–Ir atomic columns are visible in the uppermost layers of the reconstructed volume. We compare this approach to the alternative techniques of depth sectioning using differential phase contrast scanning transmission electron microscopy (DPC-STEM) and multislice ptychographic reconstruction.

47 OTHER INSTRUMENTATION↗
DOE OSTI2024

Jet-hadron correlations with respect to the event plane in $\sqrt{s_{NN}}$ = 200 GeV Au + Au collisions in STAR

Angular distributions of charged particles relative to jet axes are studied in $\sqrt{s_{NN}}$ = 200 GeV Au+Au collisions as a function of the jet orientation with respect to the event plane. This differential study tests the expected path-length dependence of energy loss experienced by a hard-scattered parton as it traverses the hot and dense medium formed in heavy-ion collisions. A second-order event plane is used in the analysis as an experimental estimate of the reaction plane formed by the collision impact parameter and the beam direction. Charged-particle jets with 15 < 𝑝 T,jet < 20 and 20 < 𝑝 T,jet < 40 ⁢GeV/𝑐 were reconstructed with the anti-𝑘 T algorithm with radius parameter setting of 𝑅 = 0.4 in the 20–50% centrality bin to maximize the initial-state eccentricity of the interaction region. The reaction plane fit method is implemented to remove the flow-modulated background with better precision than prior methods. Yields and widths of jet-associated charged-hadron distributions are extracted in three angular bins between the jet axis and the event plane. The event-plane (EP) dependence is further quantified by ratios of the associated yields in different EP bins. No dependence on orientation of the jet axis with respect to the event plane is seen within the uncertainties in the kinematic regime studied. This finding is consistent with a similar experimental observation by ALICE in $\sqrt{s_{NN}}$ = 2.76 TeV Pb-Pb collision data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

A general approach to seismic inversion with automatic differentiation

Imaging Earth structure or seismic sources from seismic data involves minimizing a target misfit function, and is commonly solved through gradient-based optimization. The adjoint-state method has been developed to compute the gradient efficiently; however, its implementation can be time-consuming and difficult. We develop a general seismic inversion framework to calculate gradients using reverse-mode automatic differentiation. The central idea is that adjoint-state methods and reverse-mode automatic differentiation are mathematically equivalent. Here, the mapping between numerical PDE simulation and deep learning allows us to build a seismic inverse modeling library, ADSeismic, based on deep learning frameworks, which supports high performance reverse-mode automatic differentiation on CPUs and GPUs. We demonstrate the performance of ADSeismic on inverse problems related to velocity model estimation, rupture imaging, earthquake location, and source time function retrieval. ADSeismic has the potential to solve a wide variety of inverse modeling applications within a unified framework.

58 GEOSCIENCES↗
DOE OSTI2022

Autoregressive Neural Network for Simulating Open Quantum Systems via a Probabilistic Formulation

The theory of open quantum systems lays the foundation for a substantial part of modern research in quantum science and engineering. Rooted in the dimensionality of their extended Hilbert spaces, the high computational complexity of simulating open quantum systems calls for the development of strategies to approximate their dynamics. In this Letter, we present an approach for tackling open quantum system dynamics. Using an exact probabilistic formulation of quantum physics based on positive operator-valued measure, we compactly represent quantum states with autoregressive neural networks; such networks bring significant algorithmic flexibility due to efficient exact sampling and tractable density. We further introduce the concept of string states to partially restore the symmetry of the autoregressive neural network and improve the description of local correlations. Efficient algorithms have been developed to simulate the dynamics of the Liouvillian superoperator using a forward-backward trapezoid method and find the steady state via a variational formulation. Our approach is benchmarked on prototypical one-dimensional and two-dimensional systems, finding results which closely track the exact solution and achieve higher accuracy than alternative approaches based on using Markov chain Monte Carlo method to sample restricted Boltzmann machines. Our Letter provides general methods for understanding quantum dynamics in various contexts, as well as techniques for solving high-dimensional probabilistic differential equations in classical setups.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Benchmarking optimization methods for materials research: Gradient descent and Bayesian optimization for lithium-ion battery aging diagnostics

Accurate and efficient parameter estimation is essential for battery diagnostics and aging analysis. Here, in this study, we compare two optimization-based approaches—gradient descent and Bayesian optimization—for extracting parameters from differential voltage analysis in lithium-ion batteries. While these techniques are widely used, their relative strengths and limitations for this application are not well understood. The study evaluates the trade-offs between these methods in terms of result quality, computational cost, and reliability within this specific application. The diagnostic results from our battery data suggest adopting gradient descent as an initial method for rapid and efficient analysis, while employing more stable optimization techniques, such as Bayesian optimization, as a verification step to mitigate potential instability. Comparing the two methods provides information on algorithmic choice, while inspiring further discussions on selecting appropriate techniques for specific research tasks.

Zhao, Ziqing [Boston Univ., MA (United States)] (O↗
DOE OSTI2023

ARKODE: A Flexible IVP Solver Infrastructure for One-step Methods

We describe the ARKODE library of one-step time integration methods for ordinary differential equation (ODE) initial-value problems (IVPs). In addition to providing standard explicit and diagonally implicit Runge–Kutta methods, ARKODE supports one-step methods designed to treat additive splittings of the IVP, including implicit-explicit (ImEx) additive Runge–Kutta methods and multirate infinitesimal (MRI) methods. We present the role of ARKODE within the SUNDIALS suite of time integration and nonlinear solver libraries, the core ARKODE infrastructure for utilities common to large classes of one-step methods, as well as its use of “time stepper” modules enabling easy incorporation of novel algorithms into the library. Numerical results show example problems of increasing complexity, highlighting the algorithmic flexibility afforded through this infrastructure, and include a larger multiphysics application leveraging multiple algorithmic features from ARKODE and SUNDIALS.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Randomized measurement protocols for lattice gauge theories

Randomized measurement protocols, including classical shadows, entanglement tomography, and randomized benchmarking are powerful techniques to estimate observables, perform state tomography, or extract the entanglement properties of quantum states. While unraveling the intricate structure of quantum states is generally difficult and resource-intensive, quantum systems in nature are often tightly constrained by symmetries. Furthermore, this can be leveraged by the symmetry-conscious randomized measurement schemes we propose, yielding clear advantages over symmetry-blind randomization such as reducing measurement costs, enabling symmetry-based error mitigation in experiments, allowing differentiated measurement of (lattice) gauge theory entanglement structure, and, potentially, the verification of topologically ordered states in existing and near-term experiments. Crucially, unlike symmetry-blind randomized measurement protocols, these latter tasks can be performed without relearning symmetries via full reconstruction of the density matrix.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2022

Development of a Single-Phase, Transient, Subchannel Code, within the MOOSE Multi-Physics Computational Framework

Subchannel codes have been widely used for thermal-hydraulics analyses in nuclear reactors. This paper details the development of a novel subchannel code within the Idaho National Laboratory’s (INL) Multi-physics Object Oriented Simulation Environment (MOOSE). MOOSE is a parallel computational framework targeted at the solution of systems of coupled, nonlinear partial differential equations, that often arise in the simulation of nuclear processes. As such, it includes codes/modules able to solve the multiple linear and nonlinear physics that describe a nuclear reactor, under normal operation conditions or accidents. This includes thermal-hydraulics, fuel performance, and neutronics codes, between others. A MOOSE-based subchannel code is a new addition to the fleet of INL-developed codes, based on the MOOSE framework. In this work, we present the derivation of the subchannel equations for a single-phase fluid, we proceed with the description of the algorithm that is used to solve these equations and describe how this algorithm was implemented within MOOSE. We also present how this code can be coupled to the BISON fuel performance code. Next, we verify the friction model and the turbulent mixing model. We calibrate the turbulent modeling parameters for momentum mixing and enthalpy mixing, C T , β. We validate the code using experimental results and last demonstrate the coupling capabilities using a simple example.

42 ENGINEERING↗
DOE OSTI2020

CSAPR2 CMAC 2.0 level c1 data.

Raw data from ARM precipitation radars must be corrected for atmospheric phenomena and instrument characteristics (e.g., attenuation, clutter) to retrieve precipitation properties. The Corrected Moments in Antenna Coordinates Version 2 (CMAC2) value-added product (VAP) is a set of algorithms and code that does such corrections, and it also retrieves precipitation quantities from the radar measurements. Similar to the X-SAPR radars at the ARM main site, C-SAPR2 radar data also needs corrections and improvements. CMAC 2.0 has been updated to work with the ARM C-SAPR 2 radar. Data and fields that have been processed to: correct for velocity aliasing, unfold and generate a cross-polarimetric phase difference that is monotonically increasing, removing impulses caused by non-uniform beam filling and phase shift on backscatter, recalculate specific differential phase using a 20-point Sobel filter on the aforementioned phase, correct for liquid path attenuation using the polarimetric signals, estimate rainfall rates at the gate using the specific attenuation. In addition, CMAC writes the data out into a community-standard format netCDF File using the CF/Radial conventions. The data are therefore compatible with new and existing National Center for Atmospheric Research (NCAR) tools such as RadXConvert for converting to a variety of popular file formats.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Technical Background and Validation Report on the Residential Water Inhalation Risk Calculator Presented in the Risk Assessment Information System

Indoor air quality (IAQ) is critical for human health. Poor IAQ is linked to respiratory issues, cardiovascular diseases, and cancer. Indoor pollutants are emitted by typical household items such as cleaning products, personal care items, building materials, and tap water - an understudied volatile organic compound (VOC) source. This document presents the Residential Water Inhalation Risk Calculator (RWIRC), which estimates daily VOC exposure concentrations from various household water uses, such as showering and dishwashing, to assess exposure risks for the most vulnerable occupant. Integrated into the Risk Assessment Information System (RAIS) and sponsored by the US Department of Energy (DOE), the calculator divides a house into three compartments: shower, bathroom, and other spaces, accounting for daily water usage patterns and calculating VOC concentrations. Exposure data generated using the calculator can assist health assessors in estimating excess lifetime cancer risk (ELCR) and hazard index (HI) from VOC inhalation. Unlike traditional exposure models that utilize Andelman’s constant, the RWIRC continuously assesses variability in VOC concentrations and environmental conditions using differential equations to track VOC concentrations and air exchange between compartments. The calculator also provides unique volatilization fractions for each chemical and appliance, enhancing accuracy of the exposure concentration estimation. The RWIRC is accessible online and allows users to customize parameters (i.e., number of bathrooms, water temperature, and exhaust fan conditions) and input VOC characteristics (i.e., tap water and ambient air concentrations). This document provides a step-by-step guide on implementing the calculator. It also provides comparisons with the ATSDR-SHOWER calculator, using eight VOCs with varying physicochemical properties to reveal differences in algorithms and output concentrations. Simulations also assess how bathroom door positions and exhaust fan usage affect VOC exposure. The calculator results can enhance EPA risk screening levels for inhalation exposure to VOCs from tap water, offering a sophisticated tool for assessing inhalation risks and improving public health protection.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Bayesian Optimization Framework for Imperfect Data or Models

Conventional Bayesian optimization methods implicitly assume that the data and model being optimized are “perfect.” This assumption leads to inaccurate posterior probability distribution functions (PDFs) when applied to “imperfect” data or models. The new Bayesian optimization framework presented in this report provides a way to parameterize the effect of imperfections usually encountered in a prior PDF of generalized data or a model on the posterior PDF. The effects of imperfections are parameterized by a set of constraints imposed on the posterior expectation values of deviations between the data and the model and on their covariance matrix elements. A particular set of values for these constraints conveys an evaluator’s best estimate of the effect of imperfections on the corresponding posterior expectation values. When a prior PDF of generalized data is assumed to be normal, an expression for a posterior PDF satisfying an arbitrary set of constraints is derived analytically for linear models. An analogous iterative algorithm is given for nonlinear models. The corresponding posterior PDF should be used to estimate any posterior expectation values in the presence of imperfections parameterized by that set of constraints. A posterior PDF of a conventional Bayesian optimization method is recovered analytically when all evaluator-specified constraints are set to zero (i.e., in the absence of any imperfections). The analytical expressions derived in this report for normal PDFs and linear models were verified numerically by a Metropolis–Hastings Monte Carlo method. The methods presented herein could be applied to any kind of data or models, including differential cross-section data or integral benchmark experiments.

97 MATHEMATICS AND COMPUTING↗