Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “inference accelerators”

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 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Optimal Power Management for Large-Scale Battery Energy Storage Systems via Bayesian Inference

Large-scale battery energy storage systems (BESS) have found ever-increasing use across industry and society to accelerate clean energy transition and improve energy supply reliability and resilience. However, their optimal power management poses significant challenges: the underlying high-dimensional nonlinear nonconvex optimization lacks computational tractability in real-world implementation, and the uncertainty of the exogenous power demand makes exact optimization difficult. This paper presents a new solution framework to address these bottlenecks. The solution pivots on introducing power-sharing ratios to specify each cell’s power quota from the output power demand. To find the optimal power-sharing ratios, we formulate a nonlinear model predictive control (NMPC) problem to achieve power-loss-minimizing BESS operation while complying with safety, cell balancing, and power supply-demand constraints. We then propose a parameterized control policy for the power-sharing ratios, which utilizes only three parameters, to reduce the computational demand in solving the NMPC problem. This policy parameterization allows us to translate the NMPC problem into a Bayesian inference problem for the sake of 1) computational tractability, and 2) overcoming the nonconvexity of the optimization problem. We leverage the ensemble Kalman inversion technique to solve the parameter estimation problem. Concurrently, a low-level control loop is developed to seamlessly integrate our proposed approach with the BESS to ensure practical implementation. This low-level controller receives the optimal power-sharing ratios, generates output power references for the cells, and maintains a balance between power supply and demand despite uncertainty in output power. We conduct extensive simulations and experiments on a 20-cell prototype to validate the proposed approach.

Battery energy storage systems (BESSs)↗
DOE OSTI2023

HD-Bind: Encoding of Molecular Structure with Low Precision, Hyperdimensional Binary Representations

Publicly available collections of drug-like molecules have grown to comprise tens of billions of compounds due to advances in combinatorial chemistry. Traditional methods for identifying "hit" molecules from a large collection of potential drug-like candidates have relied on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have the major drawback that they require exceptional computing capabilities for even relatively small collections of molecules. Hyperdimensional Computing (HDC) is a recently-proposed learning paradigm that represents data with high-dimension binary vectors; this allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas. We consider existing HDC approaches for molecular property classification and introduce two novel encodings of a commonly-used molecular representation, the extended connectivity fingerprint (ECFP). We show that HDC-based inference methods are as much as 91 times more efficient than traditional machine learning methods, and achieve an acceleration of nearly nine orders of magnitude compared to molecular docking. Our results show that HDC accelerated methods retain competitive accuracy on a number of well-studied tasks such as molecular property predictions using the MoleculeNet dataset, and bind/no-bind activity classification using the DUD-E and LIT-PCBA datasets. Our work thus motivates further investigation into molecular representation learning to develop ultraefficient pre-screening tools.

Jones, William↗
DOE OSTI2022

H-GCN: A Graph Convolutional Network Accelerator on Versal ACAP Architecture

Recently Graph Neural Networks (GNNs) have drawn tremendous attentions due to their unique capability to extend the Machine Learning (ML) approaches to broadly defined applications with unstructured data, especially graphs. Comparing with other ML modalities, the acceleration of GNNs is as critical but even more challenging due to the irregularity and heterogeneity from graph typologies that together limit the performance. Existing efforts mainly focus on handling graphs’ irregularity, however, have not studied the heterogeneity. To this end, in this work, we propose H-GCN, a PL-AIE-based hybrid accelerator that leverages the emerging heterogeneity of Xilinx Versal ACAPs to achieve high-performance GNN inference. In particular, H-GCN partitions each graph into three subgraphs based on its inherent heterogeneity and processes them using PL and the newly emerged AIE respectively. To further improve the performance, we explore the sparsity support of AIE and develop an efficient density-aware method to map tiles of SpMM onto the systolic tensor array automatically. Compared with the current state-of-the-art GCN accelerator, HGCN achieves on average 1.5× speedups.

Zhang, Chengming↗
DOE OSTI2024

Effects of catholyte aging on high-nickel NMC cathodes in sulfide all-solid-state batteries

Sulfide solid-state electrolytes (SSEs) in all-solid-state batteries (SSBs) are recognized for their high ionic conductivity and inherent safety. The LiNi 0.8 Mn 0.1 Co 0.1 O 2 (NMC811) cathode offers a high thermodynamic potential of approximately 3.8 V vs. Li/Li + and a theoretical specific capacity of 200 mA h g −1 . However, the practical utilization of NMC811 in sulfide SSBs faces significant interfacial challenges. The oxidation instability of sulfide solid electrolytes against NMC811 and the formation of the cathode electrolyte interphase (CEI) during cycling lead to degradation and reduced cell performance. Volumetric changes in NMC during lithiation and de-lithiation can also cause detachment from sulfide electrolytes or internal particle cracking. Despite extensive galvanostatic cycling studies to address the issues, the calendar life of sulfide SSBs remains poorly understood. Here, we systematically studied the effects of four different catholytes on the calendar aging of LiNbO 3 (LNO)-coated NMC811, including Li 6 PS 5 Cl (LPSCl), Li 3 InCl 6 –Li 6 PS 5 Cl (LIC–LPSCl), Li 3 YCl 6 –Li 6 PS 5 Cl (LYC–LPSCl), and Li 10 GeP 2 S 12 (LGPS). Our results indicate that LPSCl provides optimal capacity retention when stored at high state-of-charge (SOC) at room temperature, but the LIC–LPSCl cathode shows significant capacity degradation and chemical incompatibility. We also established an effective electrochemical calendar aging testing protocol to simulate daily usage, enabling quick inference of the calendar life of SSBs. In conclusion, this new testing approach accelerates materials selection strategies for high-nickel NMC composite cathodes in sulfide SSBs.

25 ENERGY STORAGE↗
DOE OSTI2025

Enhancing Gaussian Process Surrogates for Optimization and Posterior Approximation via Random Exploration

This paper proposes novel noise-free Bayesian optimization strategies that rely on a random exploration step to enhance the accuracy of Gaussian process surrogate models. The new algorithms retain the ease of implementation of the classical GP-UCB algorithm, but the additional random exploration step accelerates their convergence, nearly achieving the optimal convergence rate. Furthermore, to facilitate Bayesian inference with intractable likelihoods, we propose to utilize optimization iterates for maximum a posteriori estimation to build a Gaussian process surrogate model for the unnormalized log-posterior density. We provide bounds for the Hellinger distance between the true and the approximate posterior distributions in terms of the number of design points. We demonstrate the effectiveness of our Bayesian optimization algorithms in nonconvex benchmark objective functions, in a machine learning hyperparameter tuning problem, and in a black-box engineering design problem. The effectiveness of our posterior approximation approach is demonstrated in two Bayesian inference problems for parameters of dynamical systems.

Bayesian inference↗
DOE OSTI2026

Machine learning at the Spallation Neutron Source accelerator and target

We describe the ongoing efforts to apply Machine Learning techniques to improve the performance of our accelerator and target. Specially, we are looking to minimize halo beam losses in the absence of a proper physics model, automatically detect and log anomalies in the target support systems such as cooling, and detect and prevent errant beam pulses in the linac. We also describe the infrastructure we use to acquire and stream data to the GPU cluster for training, our code development cycle, and edge computing for model inference. To minimize halo beam losses, we use a Reinforcement Learning technique tested on a virtual accelerator. The target anomaly detection is trained on archived data using incomplete physics models and is made part of the existing target reporting system. The errant beam prevention analyzes beam current and beam phase waveforms as well as accelerator configuration data to predict errant pulses. We also develop continual learning to adapt to changes in the accelerator.

Accelerator Physics↗
NASA NTRS1985

First-order shock acceleration in solar flares

The first order Fermi shock acceleration model is compared with specific observations where electron, proton, and alpha particle spectra are available. In all events, it is found that a single shock with a compression ratio as inferred from the low energy proton spectra can reasonably produce the full proton, electron, and alpha particle spectra. The model predicts that the acceleration time to a given energy will be approximately equal for electrons and protons and, for reasonable solar parameters, can be less than 1 sec to 100 MeV.

Ellison, D. C.↗
DOE OSTI2026

Inference as a Service for HEP & NP experiments

This work presents how American Science Cloud resources and services can accelerate scientific discovery across large-scale HEP & NP experiments. Our demonstrators include use cases from intensity and energy frontiers. This is a first end-to-end multi-experiment, multi-facility demonstration of AmSC platforms.

Bhattacharya, M. [Fermilab]↗
NASA NTRS2000

Discrete and Diffuse Aurora During Varying Activity Levels: Simultaneous Fast and Polar UVI Observations

We examine simultaneous measurements of auroral electron precipitation obtained in-situ by the FAST spacecraft and remotely by Polar Ultraviolet Imagery (UVI) images for activity levels ranging from quiet to storm-time intervals. The incident energy flux measured by FAST and inferred from the UVI images agree well during quiescent periods, particularly in regions of discrete aurora in which the electron precipitation spectra are dominated by the component accelerated by a field-aligned potential. During magnetospheric substorms and active storm periods, such as those following Coronal Mass Ejection (CME) disturbances of the magnetosphere, the energy flux inferred from the UVI images generally exceeds that measured locally by FAST at the same location by as much as an order of magnitude. The auroral electrons during these active periods are dominated by diffuse precipitation which is observed up the to the highest energy channel of FAST (30 keV). These storm-time observations imply that a high energy component above 30 keV not observed by FAST may be contributing significantly to the total energy flux carried by the precipitating electrons. Observations suggest that as magnetospheric activity increases acceleration processes in the magnetosphere and pitch-angle diffusion by wave-particle interactions become more important than the ionospheric acceleration in producing the measured auroral energy fluxes.

Chua, D.↗
NASA NTRS1984

First-order Fermi shock acceleration in solar flares

The first-order Fermi shock acceleration model is compared with specific observations where electron, proton, and alpha particle spectra are available. In all events, it is found that a single shock with a compression ratio as inferred from the low-energy proton spectra can reasonably produce the full proton, electron, and alpha particle spectra. The model predicts that the acceleration time to a given energy will be approximately equal for electrons and protons and, for reasonable solar parameters, can be less than 1 sec to 100 MeV.

Ellison, D. C.↗
DOE OSTI2024

Uncertainty quantification of material parameters in modeling coupled metal and high explosive experiments

Experiments involving the coupling of metal and high explosives (HE) are of notable defense-related interest, and we seek to refine the uncertainty quantification associated with models of such experiments. In particular, our focus is on how uncertainty related to the metal constitutive model challenges our ability to infer high explosive model parameters when analyzing focused science experiments. We consider three focused experiments involving an HE accelerating metal: small plate tests with tantalum/LX-14 and tantalum/LX-17 pairings as well as a tantalum/LX-17 cylinder test. For all three models, we perform sensitivity analysis to ascertain the influence of metal strength on the coupled experimental response. Moreover, we calibrate each model in a Bayesian setting and study the quantification of metal strength on the inference of the HE parameters. Based on our results, we offer guidance for future metal/HE experiments.

36 MATERIALS SCIENCE↗
NASA NTRS1994

Transversely accelerated ions in the topside ionosphere

Data from the rocket campaigns Mechanism in the Auroral Region for Ion Energization (MARIE) and TOpside Probe of the Auroral Zone (TOPAZ) III, within regions of low-altitude transversely accelerated ions, are interpreted to explain the acceleration of the ions. Using the Monte Carlo kinetic technique to evaluate the ion heating produced by the simultaneously observed lower hybrid waves, we find that their observed electric field amplitudes are sufficient to explain the observed ion energies in the MARIE event. Much of the uncertainty in evaluating the efficiency of a plasma wave induced particle heating process which is dependent on a velocity resonance comes from the lack of information on the phase velocities of the waves. In the case of the MARIE observations, our modeling efforts show that features in the ion velocity distribution are consistent with the wave phase velocities inferred from interferometer measurements of wavelengths. The lower hybrid waves with which low-altitude transversely accelerated ions are associated are frequently observed to be concentrated in small-scale wave packets called 'spikelets'. We demonstrate through the scaling of the size of these wave packets that they are consistent with the theory of lower hybrid collapse. Using the Monte Carlo technique, we find that if the lower hybrid field energy is concentrated in these wave packets, it is still adequate to accelerate the ionospheric ions to the observed energies.

Retterer, John M.↗
DOE OSTI2022

Neural network accelerator for quantum control

Efficient quantum control is necessary for practical quantum computing implementations with current technologies. Conventional algorithms for determining optimal control parameters are computationally expensive, largely excluding them from use outside of the simulation. Existing hardware solutions structured as lookup tables are imprecise and costly. By designing a machine learning model to approximate the results of traditional tools, a more efficient method can be produced. Such a model can then be synthesized into a hardware accelerator for use in quantum systems. In this study, we demonstrate a machine learning algorithm for predicting optimal pulse parameters. This algorithm is lightweight enough to fit on a low-resource FPGA and perform inference with a latency of 175 ns and pipeline interval of 5 ns with > 0.99 gate fidelity. In the long term, such an accelerator could be used near quantum computing hardware where traditional computers cannot operate, enabling quantum control at a reasonable cost at low latencies without incurring large data bandwidths outside of the cryogenic environment.

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

Enabling Precision Neutrino Oscillation Studies with MINERvA

The next generation of neutrino experiments at accelerators aims to establish matter-antimatter asymmetry inneutrinoflavor oscillations. Precision oscillation measurements require inference of neutrino energies and flavor from the products of O(GeV) neutrino interactions on nuclei. I’ll discuss recent results from the MINERvAexperimentand how they help with this inference.

McFarland, Kevin [University of Rochester]↗
DOE OSTI2022

A Model of Double Coronal Hard X-Ray Sources in Solar Flares

A number of double coronal X-ray sources have been observed during solar flares by RHESSI, where the two sources reside at different sides of the inferred reconnection site. However, where and how these X-ray-emitting electrons are accelerated remains unclear. Here we present the first model of the double coronal hard X-ray (HXR) sources, where electrons are accelerated by a pair of termination shocks driven by bidirectional fast reconnection outflows. We model the acceleration and transport of electrons in the flare region by numerically solving the Parker transport equation using velocity and magnetic fields from the macroscopic magnetohydrodynamic simulation of a flux rope eruption. We show that electrons can be efficiently accelerated by the termination shocks and high-energy electrons mainly concentrate around the two shocks. The synthetic HXR emission images display two distinct sources extending to >100 keV below and above the reconnection region, with the upper source much fainter than the lower one. The HXR energy spectra of the two coronal sources show similar spectral slopes, consistent with the observations. Our simulation results suggest that the flare termination shock can be a promising particle acceleration mechanism in explaining the double-source nonthermal emissions in solar flares.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1979

Jupiter's magnetosphere and radiation belts

Radioastronomy and Pioneer data reveal the Jovian magnetosphere as a rotating magnetized source of relativistic particles and radio emission, comparable to astrophysical cosmic ray and radio sources, such as pulsars. According to Pioneer data, the magnetic field in the outer magnetosphere is radially extended into a highly time variable disk-shaped configuration which differs fundamentally from the earth's magnetosphere. The outer disk region, and the energetic particles confined in it, are modulated by Jupiter's 10 hr rotation period. The entire outer magnetosphere appears to change drastically on time scales of a few days to a week. In addition to its known modulation of the Jovian decametric radio bursts, Io was found to absorb some radiation belt particles and to accelerate others, and most importantly, to be a source of neutral atoms, and by inference, a heavy ion plasma which may significantly affect the hydrodynamic flow in the magnetosphere. Another important Pioneer finding is that the Jovian outer magnetosphere generates, or permits to escape, fluxes of relativistic electrons of such intensities that Jupiter may be regarded as the dominant source of 1 to 30 MeV cosmic ray electrons in the heliosphere.

Kennel, C. F.↗
NASA NTRS2012

Development of an OSSE Framework for a Global Atmospheric Data Assimilation System

Observing system simulation experiments (OSSEs) are powerful tools for estimating the usefulness of various configurations of envisioned observing systems and data assimilation techniques. Their utility stems from their being conducted in an entirely simulated context, utilizing simulated observations having simulated errors and drawn from a simulation of the earth's environment. Observations are generated by applying physically based algorithms to the simulated state, such as performed during data assimilation or using other appropriate algorithms. Adding realistic instrument plus representativeness errors, including their biases and correlations, can be critical for obtaining realistic assessments of the impact of a proposed observing system or analysis technique. If estimates of the expected accuracy of proposed observations are realistic, then the OSSE can be also used to learn how best to utilize the new information, accelerating its transition to operations once the real data are available. As with any inferences from simulations, however, it is first imperative that some baseline OSSEs are performed and well validated against corresponding results obtained with a real observing system. This talk provides an overview of, and highlights critical issues related to, the development of an OSSE framework for the tropospheric weather prediction component of the NASA GEOS-5 global atmospheric data assimilation system. The framework includes all existing observations having significant impact on short-term forecast skill. Its validity has been carefully assessed using a range of metrics that can be evaluated in both the OSSE and real contexts, including adjoint-based estimates of observation impact. A preliminary application to the Aeolus Doppler wind lidar mission, scheduled for launch by the European Space Agency in 2014, has also been investigated.

Gelaro, Ronald↗
DOE OSTI2025

DriveSense: A Noise-Resilient Framework for Driving Mode Identification

Accurate drive mode classification is essential for enhancing the reliability and predictive maintenance of heavy-duty electric trucks. This study proposes a novel fuzzy logic-based framework, DriveSense, for real-time drive mode classification, addressing key challenges such as sensor noise, transitional behaviors, and computational efficiency. The proposed approach integrates a two-stage filtering pipeline, combining adaptive outlier removal and a dynamic Kalman filter to enhance data quality. A fuzzy inference system with smoothened trapezoidal membership functions is then applied to classify driving modes into standstill, constant speed, acceleration, and deceleration while mitigating the effects of noise and edge cases. Performance evaluation using real-world and simulated drive cycles demonstrates significant improvements in classification accuracy (up to 97.8%), F1-score (up to 0.97), and robustness against noise, while reducing false positives. Comparative analysis against baseline models, demonstrates DriveSense’s superior accuracy and generalizability across diverse driving patterns. The framework’s lightweight and interpretable fuzzy inference engine operates with low computational latency, ensuring compatibility with real-time embedded systems typical of heavy-duty electric trucks. Moreover, DriveSense models transitional behaviors through overlapping fuzzy sets and adaptive borderline classification logic, enabling smooth identification of subtle shifts such as rolling stops or gradual deceleration. These results highlight DriveSense’s potential to enhance predictive maintenance strategies, reduce downtime, and support scalable, fleet-wide diagnostics.

Kumar, Praveen [Oak Ridge National Laboratory (ORN↗