Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “learning rate”

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 649 records · Page 36
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Hutchinson Trace Estimation for high-dimensional and high-order Physics-Informed Neural Networks

Physics-Informed Neural Networks (PINNs) have proven effective in solving partial differential equations (PDEs), especially when some data are available by seamlessly blending data and physics. However, extending PINNs to high-dimensional and even high-order PDEs encounters significant challenges due to the computational cost associated with automatic differentiation in the residual loss function calculation. Herein, we address the limitations of PINNs in handling high-dimensional and high-order PDEs by introducing the Hutchinson Trace Estimation (HTE) method. Starting with the second-order high-dimensional PDEs, which are ubiquitous in scientific computing, HTE is applied to transform the calculation of the entire Hessian matrix into a Hessian vector product (HVP). This approach not only alleviates the computational bottleneck via Taylor-mode automatic differentiation but also significantly reduces memory consumption from the Hessian matrix to an HVP’s scalar output. We further showcase HTE’s convergence to the original PINN loss and its unbiased behavior under specific conditions. Comparisons with the Stochastic Dimension Gradient Descent (SDGD) highlight the distinct advantages of HTE, particularly in scenarios with significant variability and variance among dimensions. We further extend the application of HTE to higher-order and higher-dimensional PDEs, specifically addressing the biharmonic equation. By employing tensor-vector products (TVP), HTE efficiently computes the colossal tensor associated with the fourth-order high-dimensional biharmonic equation, saving memory and enabling rapid computation. The effectiveness of HTE is illustrated through experimental setups, demonstrating comparable convergence rates with SDGD under memory and speed constraints. Additionally, HTE proves valuable in accelerating the Gradient-Enhanced PINN (gPINN) version as well as the Biharmonic equation. Overall, HTE opens up a new capability in scientific machine learning for tackling high-order and high-dimensional PDEs.

Curse of dimensionality↗
DOE OSTI2023

Learning-based demand-supply-coupled charging station location problem for electric vehicle demand management

We present a learning-based, demand-supply-coupled optimization model for the charging station location problem (CSLP), aiming to integrate the concept of electric vehicle (EV) charging demand management into the planning of charging infrastructures. In stage one, a gradient boosting-based learning model is developed to predict the charging demand of a charging station based on 15 defined features. Next, in stage two, a demand–supply-coupled CSLP model is developed to optimize the total charging usage rates of both existing and newly selected charging stations. We design a gradient-based stochastic spatial search algorithm to solve the proposed model. A case study with 6-year charging event data from Kansas City Missouri is performed. Results show that the proposed method can generate satisfactory charging demand predictions, and can increase charging usage rates by 14%, outperforming two benchmark approaches. Furthermore, the results of this research are poised to guide agencies in identifying optimal locations for new charging stations.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2026

Simultaneous Probe of the Charm and Bottom Quark Yukawa Couplings Using $t\bar{t}$𝐻 Events

A search for the standard model Higgs boson decaying to a charm quark-antiquark pair, 𝐻→$c\bar{c}$, produced in association with a top quark-antiquark pair ($t\bar{t}$𝐻) is presented. The search is performed with data from proton-proton collisions at √𝑠 =13 TeV, corresponding to an integrated luminosity of 138 fb−1. Advanced machine learning techniques are employed for jet flavor identification and event classification. The Higgs boson decay to a bottom quark-antiquark pair is measured simultaneously and the observed $t\bar{t}$𝐻(𝐻→$b\bar{b}$) event rate relative to the standard model expectation is 0.91$^{+0.26}_{−0.22}$. The observed (expected) upper limit on the product of production cross section and branching fraction 𝜎⁡($t\bar{t}$𝐻)⁢ℬ⁡(𝐻→$c\bar{c}$) is 0.11 (0.13) pb at 95% confidence level, corresponding to 7.8 (8.7) times the standard model prediction. When combined with the previous search for 𝐻 →$c\bar{c}$ via associated production with a 𝑊 or 𝑍 boson, the observed (expected) 95% confidence interval on the Higgs-charm Yukawa coupling modifier, 𝜅 𝑐 , is |𝜅 𝑐 | < 3.5 (2.7), the most stringent constraint to date.

Bottom quark↗
DOE OSTI2022

Group structure selection with random forests

Choosing an appropriate group structure for multigroup transport is far from an exact science. For some applications, one blindly uses a group structure developed years ago by forgotten methods. Furthermore, one sometimes uses the same group structure for a variety of problems, even if the group structure was originally developed with a certain application in mind. In this work, we create optimized group structures with simulated annealing for critical assembly test problems and apply a random forest regressor with bagging to choose the best group structure based on parameters of the different test problems. The optimized group structures were generated using a simulated annealing optimizer for several simple, spherical, and unreflected problems. The optimization was performed to minimize a cost function that included fission rate, absorption rate, leakage, and k{sub eff}. A random forest regressor was then trained on a set of International Criticality Safety Benchmark Evaluation Project inputs and used to select one of these six group structures. The trained machine learning model chose the best group structure 65% of the time, and one of the three best 89% of the time. Furthermore, it decreased the L2 error over all test problems by a factor of 25 when compared to the standard Los Alamos 70-group structure. In other words, the model chose group structure that were far more appropriate for the test problems than the traditional LANL group structure. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2020

Modeling Molten Salt Reactor Fission Product Removal with SCALE

Liquid-fuel Molten Salt Reactors (MSRs) allow for continuous online fuel treatment and processing using a variety of subsystems. Among these subsystems, the off-gas system (OGS) continuously removes fission products from the primary fuel salt through a helium sparging process. Accurately modeling fission product behavior is paramount for the development of MSRs; while there are a variety of depletion modeling and simulation tools that allow for removal of material from a defined system, most of them work in a semicontinuous, batch-wise manner. The new tools implemented into TRITON, in the SCALE suite for reactor analysis and design, aim to address this issue by implementing truly continuous material feeds and removals. In addition, material accountancy is enhanced in these tools, which enable tracking of removed material in user-specified mixtures. These continuous depletion tools provide a way to model material transport throughout systems more efficiently and accurately. This report validates the new SCALE MSR modeling capabilities using data from the Molten Salt Reactor Experiment (MSRE). After a framework was established for deriving realistic removal rates based on design specifications, these rates were used by TRITON to continuously remove specific elements while performing burnup calculations. Results were compared to those obtained in MSRE, with a focus on the 135 Xe poison fraction. After these capabilities were validated, the learned concepts and best practices were applied to model the Molten Salt Breeder Reactor (MSBR) and the Molten Salt Demonstration Reactor (MSDR) to study the effects of continuous fission product removal in fuel salt reactivity, as well as the lifetime effects.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2021

Understanding Your Utility Bills: Electricity

Electricity bills can often be difficult to decipher. Depending in the electricity provider, bills can be very long and detailed or extremely short, with only a few totals listed. Different charges can appear each month with no apparent explanation of how or why they are calculated. Learning how to read electricity bills and understand why different charges are assessed is key to maximizing cost savings through energy efficiency. This document provides guidance on the basics of understanding electricity bills and covers electricity generation, rate structures, usage trends, and savings opportunities. The information in this guide can be used to help companies collect and report their energy efficiency improvements to the DOE Better Plants Program.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2025

Efficient and Selective Chemical Transformations in Highly Charged and Confined Nanodroplets

The acceleration of chemical reaction rates and the increased product selectivity in microdroplets compared to that in bulk solutions has become a topic of increasing interest that has been extensively characterized by electrospray ionization mass spectrometry (ESI-MS). However, the sources of this acceleration and the detailed relationships between droplet properties and resulting reaction rate acceleration are still under debate. Moreover, droplet properties are governed by multiple interrelated experimental parameters, i.e., electrospray voltage, solution flow rate, etc., which makes it difficult and time-consuming to explore this diverse parameter space using traditional manual experimental or computational approaches. In this work, we developed an automated experimental platform integrating reactions in controlled charged microdroplet environments with ESI-MS characterization and sequential hybrid Bayesian modeling, as well as an optimal experimental design framework, to achieve multidimensional parameter optimization for higher reaction turnover rates, based on a model reaction of tetraethylenepentamine (TEPA) with carbon dioxide. With the current platform, we have achieved automated scans with a range of electrospray voltages and solution flow rates, and determined and optimized parameter settings to achieve increased reaction turnovers. We have also linked this platform to the underlying properties of droplets via a hybrid model incorporating physics, high-level theoretical calculations, and machine learning (ML) approaches. The autonomous platform is broadly applicable to a range of chemical reactions relevant to DOE’s mission in chemical separations, catalysis, and materials synthesis.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

L-dwarf Detection from SDSS Images using Improved Faster R-CNN

We present a data-driven approach to automatically detect L dwarfs from Sloan Digital Sky Survey (SDSS) images using an improved Faster R-CNN framework based on deep learning. The established L-dwarf automatic detection (LDAD) model distinguishes L dwarfs from other celestial objects and backgrounds in SDSS field images by learning the features of 387 SDSS images containing L dwarfs. Applying the LDAD model to the SDSS images containing 93 labeled L dwarfs in the test set, we successfully detected 83 known L dwarfs with a recall rate of 89.25% for known L dwarfs. Several techniques are implemented in the LDAD model to improve its detection performance for L dwarfs, including the deep residual network and the feature pyramid network. As a result, the LDAD model outperforms the model of the original Faster R-CNN, whose recall rate of known L dwarfs is 80.65% for the same test set. The LDAD model was applied to detect L dwarfs from a larger validation set including 843 labeled L dwarfs, resulting in a recall rate of 94.42% for known L dwarfs. The newly identified candidates include L dwarfs, late M and T dwarfs, which were estimated from color (i – z) and spectral type relation. The contamination rates for the test candidates and validation candidates are 8.60% and 9.27%, respectively. The detection results indicate that our model is effective to search for L dwarfs from astronomical images.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Efficient Data Compression for 3D Sparse TPC via Bicephalous Convolutional Autoencoder

Real-time data collection and analysis in large experimental facilities present a great challenge across multiple domains, including high energy physics, nuclear physics, and cosmology. To address this, machine learning (ML)-based methods for real-time data compression have drawn significant attention. However, unlike natural image data, such as CIFAR and ImageNet that are relatively small-sized and continuous, scientific data often come in as three-dimensional 3D data volumes at high rates with high sparsity (many zeros) and non-Gaussian value distribution. This makes direct application of popular ML compression methods, as well as conventional data compression methods, suboptimal. To address these obstacles, this work introduces a dual-head autoencoder to resolve sparsity and regression simultaneously, called Bicephalous Convolutional AutoEncoder (BCAE). This method shows advantages both in compression fidelity and ratio compared to traditional data compression methods, such as MGARD, SZ, and ZFP. To achieve similar fidelity, the best performer among the traditional methods can reach only half the compression ratio of BCAE. Moreover, a thorough ablation study of the BCAE method shows that a dedicated segmentation decoder improves the reconstruction.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Accelerating polyketide synthase engineering for high TRY production of biofuels and bioproducts (Final Technical Report)

Polyketide synthase (PKS) enzymes have a modular, deterministic logic that holds the potential to act as a flexible chemical factory for the biological production of a huge diversity of valuable small molecule compounds. However, engineering a custom PKS to produce a specific desired product currently requires years of trial and error, for reasons that remain poorly understood. In this project, we have developed a rapid, high throughput, Design-Build-Test-Learn (DBTL) cycle for polyketide synthases (PKSs) and demonstrate its utility for production of materials precursors. The objectives are 1) to develop a rapid, high-throughput (HT) DBTL cycle for PKSs that will enable production of a large number of unnatural, organic molecules on demand at high titer, rate, and yield (TRY); 2) to demonstrate the utility of the PKS DBTL cycle to produce three molecules: one commodity chemical (caprolactam or valerolactam) and two novel materials precursors (caprolactam or valerolactam derivatives); and 3) to demonstrate the utility of the PKS DBTL cycle to increase the TRY of one molecule (caprolactam or valerolactam). In this project, we have successfully demonstrated our high throughput PKS DBTL pipeline, and have biologically produced valerolactam and several other novel nylon monomers.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Quantum optimization algorithms: Energetic implications

Since the dawn of quantum computing (QC), theoretical developments like Shor's algorithm proved the conceptual superiority of QC over traditional computing. However, such quantum supremacy claims are difficult to achieve in practice because of the technical challenges of realizing noiseless qubits. In the near future, QC applications will need to rely on noisy quantum devices that offload part of their work to classical devices. One way to achieve this is by using parameterized quantum circuits in optimization or even in machine learning tasks. The energy requirements of quantum algorithms have not yet been studied extensively. Here in this article, we explore several optimization algorithms using both theoretical insights and numerical experiments to understand their impact on energy consumption. Specifically, we highlight why and how algorithms like quantum natural gradient descent, simultaneous perturbation stochastic approximations or circuit learning methods, are at least 2x to 4x more energy efficient than their classical counterparts; why feedback-based quantum optimization is energy-inefficient; and how techniques like Rosalin can improve the energy efficiency of other algorithms by a factor of ≥2 0 x. Finally, we use the NchooseK high-level programming model to run optimization problems on both gate-based quantum computers and quantum annealers. Empirical data indicate that these optimization problems run faster, have better success rates, and consume less energy on quantum annealers than on their gate-based counterparts.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Processed Lab Data for Neural Network-Based Shear Stress Level Prediction

Machine learning can be used to predict fault properties such as shear stress, friction, and time to failure using continuous records of fault zone acoustic emissions. The files are extracted features and labels from lab data (experiment p4679). The features are extracted with a non-overlapping window from the original acoustic data. The first column is the time of the window. The second and third columns are the mean and the variance of the acoustic data in this window, respectively. The 4th-11th column is the the power spectrum density ranging from low to high frequency. And the last column is the corresponding label (shear stress level). The name of the file means which driving velocity the sequence is generated from. Data were generated from laboratory friction experiments conducted with a biaxial shear apparatus. Experiments were conducted in the double direct shear configuration in which two fault zones are sheared between three rigid forcing blocks. Our samples consisted of two 5-mm-thick layers of simulated fault gouge with a nominal contact area of 10 by 10 cm^2. Gouge material consisted of soda-lime glass beads with initial particle size between 105 and 149 micrometers. Prior to shearing, we impose a constant fault normal stress of 2 MPa using a servo-controlled load-feedback mechanism and allow the sample to compact. Once the sample has reached a constant layer thickness, the central block is driven down at constant rate of 10 micrometers per second. In tandem, we collect an AE signal continuously at 4 MHz from a piezoceramic sensor embedded in a steel forcing block about 22 mm from the gouge layer The data from this experiment can be used with the deep learning algorithm to train it for future fault property prediction.

15 GEOTHERMAL ENERGY↗
DOE OSTI2024

Comparative modeling reveals the molecular determinants of aneuploidy fitness cost in a wild yeast model

Although implicated as deleterious in many organisms, aneuploidy can underlie rapid phenotypic evolution. However, aneuploidy will be maintained only if the benefit outweighs the cost, which remains incompletely understood. To quantify this cost and the molecular determinants behind it, we generated a panel of chromosome duplications in Saccharomyces cerevisiae and applied comparative modeling and molecular validation to understand aneuploidy toxicity. We show that 74%–94% of the variance in aneuploid strains’ growth rates is explained by the cumulative cost of genes on each chromosome, measured for single-gene duplications using a genomic library, along with the deleterious contribution of small nucleolar RNAs (snoRNAs) and beneficial effects of tRNAs. Machine learning to identify properties of detrimental gene duplicates provided no support for the balance hypothesis of aneuploidy toxicity and instead identified gene length as the best predictor of toxicity. Our results present a generalized framework for the cost of aneuploidy with implications for disease biology and evolution.

genic load↗
DOE OSTI2022

Bridging Python to Silicon: The SODA Toolchain

Systems performing scientific computing, data analysis, and machine learning tasks have a growing demand for application-specific accelerators that can provide high computational performance while meeting strict size and power requirements. However, the algorithms and applications that need to be accelerated are evolving at a rate that is incompatible with manual design processes based on hardware description languages. Agile hardware design tools based on compiler techniques can help by quickly producing an application-specific integrated circuit (ASIC) accelerator starting from a high-level algorithmic description. Here, we present the software-defined accelerator (SODA) synthesizer, a modular and open-source hardware compiler that provides automated end-to-end synthesis from high-level software frameworks to ASIC implementation, relying on multilevel representations to progressively lower and optimize the input code. Our approach does not require the application developer to write any register-transfer level code, and it is able to reach up to 364 giga floating point operations per second (GFLOPS)/W efficiency (32-bit precision) on typical convolutional neural network operators.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Physics-based Models, Machine Learning, and Experiment: Towards Understanding Complex Electrode Degradation

Degradation phenomena in Li-ion batteries are highly complex, coupled, and sensitive to use history and operating conditions. In this study, we show how tracking model parameters in continuum-level physics-based models, expedited by machine learning, can be useful in testing hypotheses for degradation mechanisms. An exemplary analysis using this approach is presented for a set of lithium trivanadate ( L i x V 3 O 8 ) cathodes cycled over a range of current rates. A simple cell revival process is combined with the parameter estimates over the course of cycling to extract valuable insights into cathode evolution and eliminate hypothesized degradation mechanisms for these cathodes. The presented approach is expected to be broadly applicable for degradation analysis of other electrodes.

25 ENERGY STORAGE↗
DOE OSTI2026

Characterization and automated optimization of laser-driven proton beams from converging liquid sheet jet targets

Compact, stable, and versatile laser-driven ion sources hold great promise for applications ranging from medicine to materials science and fundamental physics. While single-shot sources have demonstrated favorable beam properties, including the peak fluxes necessary for several applications, high-repetition-rate operation will be necessary to generate and sustain the high average flux needed for many of the most exciting applications of laser-driven ion sources. Further, to navigate through the high-dimensional space of laser and target parameters toward experimental optima, it is essential to develop ion acceleration platforms compatible with machine learning techniques and capable of autonomous real-time optimization. Here, we present a multi-Hz ion acceleration platform employing a liquid sheet jet target. We characterize the laser-plasma interaction and the laser-driven proton beam across a variety of key parameters governing the interaction using an extensive suite of online diagnostics. We also demonstrate real-time, closed-loop optimization of the ion beam maximum energy by tuning the laser wave front using a Bayesian optimization scheme. This approach increased the maximum proton energy by 11% compared to a manually optimized wave front by enhancing the energy concentration within the laser focal spot, demonstrating the potential for closed-loop optimization schemes to tune future ion accelerators for robust high-repetition-rate operation.

Glenn, G. D. [SLAC National Accelerator Laboratory↗
DOE OSTI2022

A Machine Learning-Assisted Framework to Control Thermally Anisotropic Building Envelopes in Residential Buildings

To curb the energy consumption of buildings and their related CO2 emissions, Oak Ridge National Laboratory (ORNL) has developed the thermally anisotropic building envelope (TABE) —a multi-layer design comprising insulation materials and metal foils connected to thermal loops. In this study, a machine learning-assisted framework was developed to control the TABE in residential buildings to reduce the computation load for future optimal rule-based control and application. First, a 2D finite element model was established in COMSOL to calculate the hourly heat flux through exterior walls installed with the TABE. Then, TABE wall heat fluxes were simulated for various given indoor and outdoor boundary conditions, as well as thermal loops fluid temperatures and flow rates. Since the finite element simulations are computationally expensive, an artificial neural network (ANN) was then trained to use as a proxy of the finite element (COMSOL) modeling. Finally, the trained ANN model was coupled with the EnergyPlus model to predict the energy consumption of a US Department of Energy prototype single-family house installed with the TABE. An optimal simple rule-based control was determined from predefined rules for a case study. The results demonstrate that the developed machine learning–assisted framework can reduce 99.9% of the computation time while efficiently managing residential building energy for installed TABE walls.

Shen, Zhenglai↗
DOE OSTI2026

Multi‐Scale Model‐Informed Deep Learning for Plasma‐Nanoparticle Interaction

The Overarching Goal of this proposed research is to understand and quantitively determine the interactions between non-thermal plasma (hot electrons, reactive radicals, vibrationally excited species) and surface reactions on influencing the activity and selectivity of the desired reactions via developing multi-scale model informed deep learning algorithm. Investigating non-thermal plasma-surface interaction is feasible due to the low bulk temperature in the discharge region. To investigate the role of plasma-nanoparticle interaction on enhancing the reaction kinetics, we will focus on ammonia cracking to generate clean hydrogen over earth-abundant, non-critical metallic nanoparticles, which is of great significance for decarbonization. We hypothesize that (1) reactive radicals interacting with surface reaction species via Eley–Rideal mechanism will significantly lower the energetics of the potential rate-limiting step of nitrogen formation; (2) the surface will be charged heterogeneously under non-thermal plasma conditions and the charged site will lower the energetics of ammonia cracking through Langmuir– Hinshelwood mechanism; (3) vibrationally excited ammonia will further promote the initial N-H bond cleavage. To access the hypothesis, we will (1) reveal the surface charge effects on tunning the reaction energetics via interpretable, physics-informed deep learning accelerated density functional theory (DFT) calculations; (2) determine the reactive radicals interacting with surface reaction species on tuning the reaction energetics via DFT; (3) reveal the surface charge effects on tunning the reaction energetics via DFT and deep learning models, (4) quantify how vibrationally excited species, reactive radicals, and surface charging effects on enhancing the catalysis via developing DFT-based microkinetic modeling (MKM) and active learning. Deep and active learning of plasma-nanoparticle interactions effects on enhancing ammonia cracking to generate hydrogen represents a new paradigm for designing high performance plasma materials. The fundamental science of how plasma-nanoparticle interactions will change the plasma kinetics and will improve the energy efficiency for decarbonization and sustainability. The interpretable and physics-informed machine learning model will accelerate low temperature plasma chemistry and material discovery with physics rules and model interpretation.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗