Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “small dataset”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Deep Learning Explicit Differentiable Predictive Control Laws for Buildings

We present a differentiable predictive control (DPC) methodology for learning constrained control laws for unknown nonlinear systems. DPC poses an approximate solution to multiparametric programming problems emerging from explicit nonlinear model predictive control (MPC). Contrary to approximate MPC, DPC does not require supervision by an expert controller. Instead, a system dynamics model is learned from a small dataset of recorded observations of the perturbed system's dynamics and the control law is optimized offline by interaction with the learned system model. The DPC method is based on two sequential steps, i) system identification using a constrained neural state-space model, and ii) optimization of an explicit control law parametrized by another neural network in closed-loop simulation with the identified neural state-space model. The combination of a differentiable closed-loop system and penalty methods for constraint handling of system outputs and inputs allows us to optimize the control law's parameters directly by backpropagating economic MPC loss through the learned system model. By incorporating domain knowledge and leveraging established techniques from optimal control, our method leverages deep neural networks as nonlinear function approximators for system identification and control while avoiding concomitant costs of intractably large datasets, and computationally expensive over-parametrized models. The scalability, data efficiency, and constrained optimal control capability of the proposed DPC method are demonstrated in simulation using a multi-zone building emulator.

Drgona, Jan↗
DOE OSTI2024

Hierarchical Bayesian modeling for Inverse Uncertainty Quantification of system thermal-hydraulics code using critical flow experimental data

The best estimate plus uncertainty methodology in nuclear system thermal-hydraulic studies necessitates a comprehensive understanding of uncertainties in system code predictions. The forward uncertainty quantification (UQ) process involves the propagation of input uncertainties through the computational models to obtain uncertainties in the outputs. To this end, achieving an accurate estimation of input uncertainties is important, which is the focus of inverse UQ (IUQ). Traditionally, research in Bayesian IUQ within the nuclear engineering domain has largely relied on single-level Bayesian inference. While being effective for relatively small datasets, this approach encounters limitations for cases with large datasets. The use of a single-level model may prove inefficient, as the resultant posterior distributions can significantly differ when distinct subsets of data are employed. To address this issue, we employ an hierarchical Bayesian model for IUQ. Furthermore, this approach involves organizing observations into different groups based on the test conditions, thereby accommodating varying calibration parameters across these distinct groups. In this study, we developed and implemented a hierarchical Bayesian IUQ method to consider the grouping effect of critical flow measurement data from various geometries. Comparing the outcomes of IUQ under different selections of test data using hierarchical Bayesian IUQ against those obtained from single-level Bayesian IUQ, the forward propagation of hierarchical Bayesian IUQ results demonstrates a notably improved agreement with the experimental data.

42 ENGINEERING↗
DOE OSTI2024

Machine learning for the redox potential prediction of molecules in organic redox flow battery

Here, organic redox flow batteries (ORFB) are recognized as an innovative technology for the large-scale storage of renewable energy. The redox potential of organic redox-active molecules plays a vital role in their performance. Advanced screening techniques like high-throughput experiment and machine learning (ML) have significantly enhanced organic material performance and transformed the field of ORFB. However, the scarcity of experimental data poses a considerable challenge for ML model development in this domain. In our study, we developed lightweight graph-based Gaussian process regression (GPR) models with GPU-accelerated marginalized graph kernel and hybrid kernel to predict the redox potentials of organic redox-active molecules for ORFBs, specifically focusing on small datasets. To evaluate model accuracy, we created a new experimental database of organic redox-active molecules by the data from hundreds of published papers and assembled previous computational datasets. We also considered some key parameters, such as pH conditions and solvent type, to assess their impact on redox potential prediction. Our GPR model predicted redox potentials with high accuracy across all datasets using minimal training data. The study provides powerful tools for molecule screening and design and delivers valuable guidance on designing training datasets for costly experiments.

25 ENERGY STORAGE↗
DOE OSTI2025

Graph neural networks for mechanical property prediction of 2D fiber composites

This work investigates the ability of graph neural networks (GNNs) to homogenize 2D fiber composite microstructures. We use different inhomogeneity and anisotropy indices to motivate and show that the Volume Elements (VEs) used in ML methods should ideally be far from their Representative Volume Element (RVE) size limit and, consequently, are notably anisotropic. Hence, training only the isotropic limit properties may not be acceptable. Another aspect is the need to normalize elastic stiffness values for ML, especially when high elastic contrast ratios are encountered between composite phases or in the material set. We introduce a normalization technique based on the mean-field method (MFM) to handle such high contrast ratios and train for the entire stiffness tensor. We show that the proposed GNN approaches exhibit high accuracy and efficiency compared to traditional methods and convolutional neural networks, utilizing unstructured graphs constructed from microstructure topology. Our model successfully predicts the stiffness tensor, peak strength under bulk damage, and brittle fracture initiation strength across diverse microstructure configurations while maintaining high accuracy even for extreme material contrasts and volume fractions. We also present a method to improve prediction accuracy for small dataset sizes using Voronoi partitioning.

Brittle strength↗
DOE OSTI2024

Li-ion battery design through microstructural optimization using generative AI

Lithium-ion batteries are used across various applications, necessitating tailored cell designs to enhance performance. Optimizing electrode manufacturing parameters is a key route to achieving this, as these parameters directly influence the microstructure and performance of the cells. However, linking process parameters to performance is complex, and experimental or modeling campaigns are often slow and expensive. This study introduces a fast computational optimization framework for electrode manufacturing parameters. A generative model, trained on a small dataset of microstructural images associated with different manufacturing parameters, efficiently generates representative microstructures for new parameters. This model is integrated into a Bayesian optimization loop that includes microstructure generation, characterization, and simulation, aiming to find optimal manufacturing parameters for a particular application. Significant improvement in the energy density of a 4680 cell is achieved through bespoke cell design, highlighting the importance of cell-scale normalization. The framework’s modularity allows its application to various advanced materials manufacturing scenarios.

batteries↗
DOE OSTI2022

Rapid macrovoid characterization in membranes prepared via nonsolvent-induced phase separation: A comparison between 2D and 3D techniques

Optimizing the performance of asymmetric membranes prepared via nonsolvent-induced phase separation (NIPS) requires a quantitative understanding of how processing variables influence membrane morphology. Presently, the most useful structural quantification techniques require 3D visualization of the membrane structure and are best suited for studies seeking detailed information on small datasets. This study proposes and validates a rapid and accurate technique for quantifying macroporosity (i.e., D m ), a simple descriptor of sublayer macrovoid content in asymmetric membranes D m . values measured from segmented cross-sectional imaging performed via X-ray computed tomography (XCT) and scanning electron microscopy (SEM) are presented and compared for three asymmetric membranes prepared from commercial polymers. Importantly, analyses of 3D XCT membrane reconstructions reveal that D m is described by a single, centralized mean, which demonstrates that macrovoid content is spatially homogenous. Thus, D m can be approximated from limited sampling of the 2D cross-sectional membrane structure via SEM. A proposed 2D SEM sampling method provides D m estimates within ±6% of corresponding 3D XCT values with 30 independent measurements for the three membranes. Further sensitivity is achieved using complementary descriptors such as macrovoid count density (i.e., C m ). This technique is thus a useful tool for characterizing macroporosity from a broad selection of membrane samples.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2017

Quantitative decoding of the response a ceramic mixed potential sensor array for engine emissions control and diagnostics

The development of on-board sensors for emissions monitoring is necessary for continuous monitoring of the performance of catalytic systems in automobiles. We have fabricated mixed potential electrochemical gas sensing devices with Pt, La 0.8 Sr 0.2 CrO 3 (LSCO), and Au/Pd alloy electrodes and a porous yttria-stabilized zirconia electrolyte. The three-electrode design takes advantage of the preferential selectivity of the Pt + Au/Pd and Pt + LSCO pairs towards different species of gases and has additional tunable selectivity achieved by applying a current bias to the latter pair. Voltages were recorded in single, binary, and ternary gas streams of NO, NO 2 , C 3 H 8 , and CO. We have also trained artificial neural networks to examine the voltage output from sensors in biased and unbiased modes to both identify which single test gas or binary mixture of two test gases is present in a gas stream as well as extract concentration values. We were then able to identify single and binary mixtures of these gases with accuracy of at least 98%. For determining concentration, the peak in the error distribution for binary mixtures was 5% and 80% of test data fell under <12% error. The sensor stability was also evaluated over the course of over 100 days and the ability to retrain ANNs with a small dataset was demonstrated.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2025

Deep Learning with Reflection High-Energy Electron Diffraction Images to Predict Cation Ratio in Sr 2 x Ti 2(1– x ) O 3 Thin Films

Machine learning (ML) with in-situ diagnostics offers a transformative approach to accelerate, understand, and control thin film synthesis by uncovering relationships between synthesis conditions and material properties. In this study, we demonstrate the application of deep learning to predict the stoichiometry of Sr 2x Ti 2(1–x) O 3 thin films using reflection high-energy electron diffraction images acquired during pulsed laser deposition. A gated convolutional neural network trained for regression of the Sr atomic fraction achieved accurate predictions with a small dataset of 31 samples. Explainable AI techniques revealed a previously unknown correlation between diffraction streak features and cation stoichiometry in Sr 2x Ti 2(1–x) O 3 thin films. Here, our results demonstrate how ML can be used to transform a ubiquitous in-situ diagnostic tool, that is usually limited to qualitative assessments, into a quantitative surrogate measurement of continuously valued thin film properties. Such methods are critically needed to enable real-time control, autonomous workflows, and accelerate traditional synthesis approaches.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2024

Reparameterizing Litter Decomposition Using a Simplified Monte Carlo Method Improves Litter Decay Simulated by a Microbial Model and Alters Bioenergy Soil Carbon Estimates

Litter decomposition determines soil organic matter (SOM) formation and plant-available nutrient cycles. Therefore, accurate model representation of litter decomposition is critical to improving soil carbon (C) projections of bioenergy feedstocks. Soil C models that simulate microbial physiology (i.e., microbial models) are new to bioenergy agriculture, and their parameterization is often based on small datasets or manual calibration to reach benchmarks. Here, we reparameterized litter decomposition in a microbial soil C model (CORPSE - Carbon, Organisms, Rhizosphere, and Protection in the Soil Environment) using the continental-scale Long-term Inter-site Decomposition Experiment Team (LIDET) dataset which documents decomposition across a range of litter qualities over a decade. Here, we conducted a simplified Monte Carlo simulation that constrained parameter values to reduce computational costs. The LIDET-derived parameters improved modeled C and nitrogen (N) remaining, decomposition rates, and litter mean residence times as compared to Baseline parameters. We applied the LIDET litter decomposition parameters to a microbial bioenergy model (Fixation and Uptake of Nitrogen – Bioenergy Carbon, Rhizosphere, Organisms, and Protection) to examine soil C estimates generated by Baseline and LIDET parameters. LIDET parameters increased estimated soil C in bioenergy feedstocks, with even greater increases under elevated plant inputs (i.e., by increasing residue, N fertilization). This was due to the integrated effects of plant litter quantity, quality, and agricultural practices (tillage, fertilization). Collectively, we developed a simple framework for using large-scale datasets to inform the parameterization of microbial models that impacts projections of soil C for bioenergy feedstocks.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2013

Automated fit of high-dimensional potential energy surfaces using cluster analysis and interpolation over descriptors of chemical environment

We present a method for fitting high-dimensional potential energy surfaces that is almost fully automated, can be applied to systems with various chemical compositions, and involves no particular choice of function form. We tested it on four systems: Ag{sub 20}, Sn{sub 6}Pb{sub 6}, Si{sub 10}, and Li{sub 8}. The cost for energy evaluation is smaller than the cost of a density functional theory (DFT) energy evaluation by a factor of 1500 for Li{sub 8}, and 60 000 for Ag{sub 20}. We achieved intermediate accuracy (errors of 0.4 to 0.8 eV on atomization energies, or, 1% to 3% on cohesive energies) with rather small datasets (between 240 and 1400 configurations). We demonstrate that this accuracy is sufficient to correctly screen the configurations with lowest DFT energy, making this function potentially very useful in a hybrid global optimization strategy. We show that, as expected, the accuracy of the function improves with an increase in the size of the fitting dataset.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Review and meta-analysis of electron temperatures from high-intensity laser–solid interactions

The accelerated electron spectrum from high-intensity laser–solid interaction is often conveniently described using a Boltzmann distribution, whose temperature is known within the field as the hot-electron temperature. The importance of the electron temperature is highlighted by the sheer number of experimental and simulation studies on the subject over the past three decades. Recently, multi-kJ, multi-ps pulses have yielded electron spectra with temperatures far beyond the expected ponderomotive result. Expressions that predict the electron temperature considering laser parameters beyond intensity and wavelength have been developed, albeit using small datasets. In this review, we present what is, to the best of our knowledge, the largest dataset of electron temperatures gathered from experimental measurements and particle-in-cell simulations. This dataset allows us to compare existing analytical and empirical hot-electron temperature scaling models over a wide parameter range. We also develop new scaling models that incorporate the laser pulse duration of the laser and the plasma scale length. Three models that include pulse-duration and scale length dependence are especially successful at predicting both simulated and experimental data. The dataset will soon be made publicly available to encourage further investigation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

Stochastic machine learning via sigma profiles to build a digital chemical space

This work establishes a different paradigm on digital molecular spaces and their efficient navigation by exploiting sigma profiles. To do so, the remarkable capability of Gaussian processes (GPs), a type of stochastic machine learning model, to correlate and predict physicochemical properties from sigma profiles is demonstrated, outperforming state-of-the-art neural networks previously published. The amount of chemical information encoded in sigma profiles eases the learning burden of machine learning models, permitting the training of GPs on small datasets which, due to their negligible computational cost and ease of implementation, are ideal models to be combined with optimization tools such as gradient search or Bayesian optimization (BO). Gradient search is used to efficiently navigate the sigma profile digital space, quickly converging to local extrema of target physicochemical properties. While this requires the availability of pretrained GP models on existing datasets, such limitations are eliminated with the implementation of BO, which can find global extrema with a limited number of iterations. A remarkable example of this is that of BO toward boiling temperature optimization. Holding no knowledge of chemistry except for the sigma profile and boiling temperature of carbon monoxide (the worst possible initial guess), BO finds the global maximum of the available boiling temperature dataset (over 1,000 molecules encompassing more than 40 families of organic and inorganic compounds) in just 15 iterations (i.e., 15 property measurements), cementing sigma profiles as a powerful digital chemical space for molecular optimization and discovery, particularly when little to no experimental data is initially available.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Enabling robust offline active learning for machine learning potentials using simple physics-based priors

Abstract Machine learning surrogate models for quantum mechanical simulations have enabled the field to efficiently and accurately study material and molecular systems. Developed models typically rely on a substantial amount of data to make reliable predictions of the potential energy landscape or careful active learning (AL) and uncertainty estimates. When starting with small datasets, convergence of AL approaches is a major outstanding challenge which has limited most demonstrations to online AL. In this work we demonstrate a Δ-machine learning (ML) approach that enables stable convergence in offline AL strategies by avoiding unphysical configurations with initial datasets as little as a single data point. We demonstrate our framework’s capabilities on a structural relaxation, transition state calculation, and molecular dynamics simulation, with the number of first principle calculations being cut down anywhere from 70%–90%. The approach is incorporated and developed alongside AMP torch , an open-source ML potential package, along with interactive Google Colab notebook examples.

Shuaibi, Muhammed↗
DOE OSTI2024

Picasso: Memory-Efficient Graph Coloring Using Palettes With Applications in Quantum Computing

A coloring of a graph is an assignment of colors to vertices such that no two neighboring vertices have the same color. The need for memory-efficient coloring algorithms is motivated by their application in computing clique partitions of graphs arising in quantum computations where the objective is to map a large set of Pauli strings into a compact set of unitaries. We present Picasso, a randomized memory-efficient iterative parallel graph coloring algorithm with theoretical sublinear space guarantees under practical assumptions. The parameters of our algorithm provide a trade-off between coloring quality and resource consumption. To assist the user, we also propose a machine learning model to predict the coloring algorithm’s parameters considering these trade-offs. We provide a sequential and a parallel implementation of the proposed algorithm. We perform an experimental evaluation on a 64-core AMD CPU equipped with 512 GB of memory and an Nvidia A100 GPU with 40GB of memory. For a small dataset where existing coloring algorithms can be executed within the 512 GB memory budget, we show up to 68× memory savings. On massive datasets we demonstrate that GPU-accelerated Picasso can process inputs with 49.5× more Pauli strings (vertex set in our graph) and 2,478× more edges than state-of-the-art parallel approaches.

artificial intelligence, quantum computing↗
DOE OSTI2020

The Mertens Unrolled Network (MU-Net): A High Dynamic Range Fusion Neural Network for Through the Windshield Driver Recognition

Face recognition of vehicle occupants through windshields in unconstrained environments poses a number of unique challenges ranging from glare, poor illumination, driver pose and motion blur. In this paper, we further develop the hardware and software components of a custom vehicle imaging system to better overcome these challenges. After the build out of a physical prototype system that performs High Dynamic Range (HDR) imaging, we collect a small dataset of through-windshield image captures of known drivers. We then reformulate the classical Mertens-Kautz-Van Reeth HDR fusion algorithm as a pre-initialized neural network, which we name the Mertens Unrolled Network (MU-Net), for the purpose of fine-tuning the HDR output of through-windshield images. Reconstructed faces from this novel HDR method are then evaluated and compared against other traditional and experimental HDR methods in a pre-trained state-of-the-art (SOTA) facial recognition pipeline, verifying the efficacy of our approach.

Ruby, Max↗
DOE OSTI2025

ChatHPC: Building the Foundations for a Productive and Trustworthy AI-Assisted HPC Ecosystem

ChatHPC democratizes large language models for the high-performance computing (HPC) community by providing the infrastructure, ecosystem, and knowledge needed to apply modern generative AI technologies to rapidly create specific capabilities for critical HPC components while using relatively modest computational resources. Our divide-and-conquer approach focuses on creating a collection of reliable, highly specialized, and optimized AI assistants for HPC based on the cost-effective and fast Code Llama fine-tuning processes and expert supervision. We target major components of the HPC software stack, including programming models, runtimes, I/O, tooling, and math libraries. Thanks to AI, ChatHPC provides a more productive HPC ecosystem by boosting important tasks related to portability, parallelization, optimization, scalability, and instrumentation, among others. With relatively small datasets (on the order of KB), the AI assistants, which are created in a few minutes by using one node with two NVIDIA H100 GPUs and the ChatHPC library, can create new capabilities with Meta’s 7-billion parameter Code Llama base model to produce high-quality software with a level of trustworthiness of up to 90% higher than the 1.8-trillion parameter OpenAI ChatGPT-4o model for critical programming tasks in the HPC software stack.

Young, Aaron [ORNL] (ORCID:0000000254484667)↗
DOE OSTI2023

Crystal Diffraction Prediction and Partiality estimation using Gaussian basis functions

A small dataset with good data quality, useful as a benchmark. There exists a two-fold indexing ambiguity, so when processing with CrystFEL you will need to do: ambigator -o disambiguated.stream -y m-3 -w m-3m -j 72 ambiguous.stream Please check the README file for more information about the dataset. ----- Begin unit cell ----- CrystFEL unit cell file version 1.0 lattice_type = cubic centering = I a = 103.40 A b = 103.40 A c = 103.40 A al = 90.00 deg be = 90.00 deg ga = 90.00 deg ; Please note: this is the target unit cell. ; The actual unit cells produced by indexing depend on many other factors. ----- End unit cell -----

Cydia pomonella granulovirus hull protein↗
DOE OSTI2023

Integrated Design of Ultradurable, Low CO 2 Alternative Binder Systems via Machine Learning

This ARPA-E project developed a machine learning tool to use in formulation design of cementitious binders for concrete having 50% less embodied CO 2 and possessing twice the durability compared to concrete based on ordinary portland cement (OPC) binders. The technical focus was on limestone/calcined clay cement (LC3), the leading replacement for OPC. Here, hierarchical machine learning (HML) was used to model the flowability, set time, strength, and durability of LC3 concrete. This methodology identifies latent variables derived from domain knowledge and empirical models that develop an accurate model for a response surface from small datasets. For the flowability metric, particle packing was a dominant factor, while strength and durability were both strongly determined by the fraction of metakaolin and the water:solids ratio. Under constraints of water:binder ratio, material performance metrics, embodied CO 2 , and cost per tonne of OPC, multi-objective optimization was used to design binders parameterized by the mineral composition replacing OPC, particle size distributions, and water:solids ratio. The trained algorithm was able to predict multiple mixes met these performance criteria, and experimental testing validated the predictions. The model demonstrated here is relevant for North America, where pure kaolin deposits are found broadly. The approach is being taken forward into commercial application by Ansatz AI, a materials informatics company founded by PI Washburn and co-PI Poczos. Through collaborations with the cement and concrete industry, and funding from SBIR programs, a commercial software will be developed in future research.

36 MATERIALS SCIENCE↗