Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “factorization machine”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
DOE OSTI2015

Design Concepts for an Outage Control Center Information Dashboard

The nuclear industry, and the business world in general, is facing a rapidly increasing amount of data to be dealt with on a daily basis. In the last two decades, the steady improvement of data storage devices and means to create and collect data along the way influenced the manner in which we deal with information. Most data is still stored without filtering and refinement for later use. Many functions at a nuclear power plant generate vast amounts of data, with scheduled and unscheduled outages being a prime example of a source of some of the most complex data sets at the plant. To make matters worse, modern information and communications technology is making it possible to collect and store data faster than our ability to use it for making decisions. However, in most applications, especially outages, raw data has no value in itself; instead, managers, engineers and other specialists want to extract the information contained in it. The complexity and sheer volume of data could lead to information overload, resulting in getting lost in data that may be irrelevant to the task at hand, processed in an inappropriate way, or presented in an ineffective way. To prevent information overload, many data sources are ignored so production opportunities are lost because utilities lack the ability to deal with the enormous data volumes properly. Decision-makers are often confronted with large amounts of disparate, conflicting and dynamic information, which are available from multiple heterogeneous sources. Information and communication technologies alone will not solve this problem. Utilities need effective methods to exploit and use the hidden opportunities and knowledge residing in unexplored data resources. Superior performance before, during and after outages depends upon the right information being available at the right time to the right people. Acquisition of raw data is the easy part; instead, it is the ability to use advanced analytical, data processing and data visualization methods to turn the data into reliable information and comprehensible, actionable information. Techniques like data mining, filtering and analysis only work reliably for well-defined and well-understood problems. The path from data to decision is more complex. The ability to communicate knowledge during outages and emergent issues is crucial. This paper presents an approach to turn the unused data into an opportunity: applying principles from semiotics, human factors and visual analytics to transform the traditional way of processing outage data into media that will improve the collective situation awareness, knowledge, decisions, actions and overall performance of the entire outage team, and also support the reliability, quality and overall effectiveness of maintenance work. The application of the proposed visualization methods will become the medium of a semi-automated analytical process where humans and machines cooperate using their respective, distinct capabilities for the most effective results.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2023

MACHINE LEARNING-ENABLED PREDICTION OF TRANSIENT INJECTION MAP IN AUTOMOTIVE INJECTORS WITH UNCERTAINTY QUANTIFICATION

Accurate prediction of injection profiles is a critical aspect of linking injector operation with engine performance and emissions. However, highly resolved injector simulations can take one to two weeks of wall-clock time, which is incompatible with engine design cycles with desired turnaround times of less than a day. Hence, it is important to reduce the time-to-solution of the internal flow simulations by several orders of magnitude to make it compatible with engine simulations. This work demonstrates a data-driven approach for tackling the computational overhead of injector simulations, whereby the transient injection profiles are emulated for a side-oriented, single-hole diesel injector using a Bayesian machine-learning framework. First, an interpretable Bayesian learning strategy was employed to understand the effect of design parameters on the total void fraction field. Then, autoencoders are utilized for efficient dimensionality reduction of the flowfields. Gaussian process models are finally used to predict the spatiotemporal void fraction field at the injector exit for unknown operating conditions. The Gaussian process models produce principled uncertainty estimates associated with the emulated flowfields, which provide the engine designer with valuable information of where the data-driven predictions can be trusted in the design space. The Bayesian flowfield predictions are compared with the corresponding predictions from a deep neural network, which has been transfer-learned from static needle simulations from a previous work by the authors. The emulation framework can predict the void fraction field at the exit of the orifice within a few seconds, thus achieving a speed-up factor of up to 38 x 10(6) over the traditional simulation-based approach of generating transient injection maps.

machine learning↗
DOE OSTI2023

Electric and Hydraulic Propel Torque Modulation for a Compact Track Loader With the Hybrid Hydraulic Electric Architecture (HHEA)

Abstract The Hybrid Hydraulic Electric Architecture (HHEA) has previously been proposed for off-highway vehicles to reap the efficiency and controllability benefits of electrification without needing very large electric motors. This is achieved with the use of a set of selectable common pressure rails to transmit the majority of power and small electric motors to modulate that power. Previous work has shown significant energy savings for the work circuits of a variety of machines. In this paper, the energy saving potential of HHEA for the propel circuit of a compact track loader is studied. The ports of the track hydraulic motors are selectably connected to the common pressure rails, and instead of using the electric assist motors to buck/boost pressure, as in HHEA for linear actuators, small electric assist motors are used to add/subtract torque directly. The interplay between the torque limits of the electric motors and the ability of the hydraulic motor to vary displacements is studied, along with the effect these factors have on energy saving potential. It is found that the ability to vary the displacement of the hydraulic motor allows for: more efficient electric motor operating conditions, reduced electric torque requirement, and reduced pressure rail switching events. All three of these advantages can be achieved at once using variable displacements; but trade-offs exist between these advantages (i.e. improved efficiency can be achieved at the expense of a larger electric torque requirement). Overall, the HHEA can reduce energy consumption by ∼ 36% compared to the stock machine, depending on the hydraulic motor’s ability to vary displacements, and assuming the electric motor torque is limited to 20% of that required in a direct electrification scheme.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2017

Ensuring operator reliability in digital control rooms - 102

The U.S. fleet of nuclear power plants is aging, and efforts are being made to modernize instrumentation and control systems including integration of digital human-machine interfaces (HMIs) into the control room. Automation and control technologies have advanced since original systems were installed in the plants. This presents unique opportunities and challenges in ensuring the replacements are a benefit to the operators and the overall safe and reliable operation of the plant. We discuss trends we have observed from assisting utilities in their modernization efforts. Automation and evolving control rooms come with their own host of problems that existing human factors and engineering methods fail to capture. Traditional approaches are analytical approaches to ensuring reliability. Here we discuss some of the shortfalls of these traditional approaches as plants become increasingly more complex. We suggest that ensuring reliability for modernized and next generation control room will require expanding the scope of what was traditionally considered human factors as well as exploring novel approaches and methodologies for dealing with complexity and rapidly advancing technology. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2016

Improving Lidar Turbulence Estimates for Wind Energy

Remote sensing devices (e.g., lidars) are quickly becoming a cost-effective and reliable alternative to meteorological towers for wind energy applications. Although lidars can measure mean wind speeds accurately, these devices measure different values of turbulence intensity (TI) than an instrument on a tower. In response to these issues, a lidar TI error reduction model was recently developed for commercially available lidars. The TI error model first applies physics-based corrections to the lidar measurements, then uses machine-learning techniques to further reduce errors in lidar TI estimates. The model was tested at two sites in the Southern Plains where vertically profiling lidars were collocated with meteorological towers. This presentation primarily focuses on the physics-based corrections, which include corrections for instrument noise, volume averaging, and variance contamination. As different factors affect TI under different stability conditions, the combination of physical corrections applied in L-TERRA changes depending on the atmospheric stability during each 10-minute time period. This stability-dependent version of L-TERRA performed well at both sites, reducing TI error and bringing lidar TI estimates closer to estimates from instruments on towers. However, there is still scatter evident in the lidar TI estimates, indicating that there are physics that are not being captured in the current version of L-TERRA. Two options are discussed for modeling the remainder of the TI error physics in L-TERRA: machine learning and lidar simulations. Lidar simulations appear to be a better approach, as they can help improve understanding of atmospheric effects on TI error and do not require a large training data set.

17 WIND ENERGY↗
DOE OSTI2020

Fast Scanning Probe Microscopy via Machine Learning: Non-Rectangular Scans with Compressed Sensing and Gaussian Process Optimization

Fast scanning probe microscopy enabled via machine learning allows for a broad range of nanoscale, temporally resolved physics to be uncovered. However, such examples for functional imaging are few in number. Here, using piezoresponse force microscopy (PFM) as a model application, a factor of 5.8 reduction in data collection using a combination of sparse spiral scanning with compressive sensing and Gaussian process regression reconstruction is demonstrated. It is found that even extremely sparse spiral scans offer strong reconstructions with less than 6% error for Gaussian process regression reconstructions. Further, the error associated with each reconstructive technique per reconstruction iteration is analyzed, finding the error is similar past ≈15 iterations, while at initial iterations Gaussian process regression outperforms compressive sensing. Finally, this study highlights the capabilities of reconstruction techniques when applied to sparse data, particularly sparse spiral PFM scans, with broad applications in scanning probe and electron microscopies.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Dynamics and lipid membrane coupling of the RAS-RAF complex revealed via multiscale simulations

To gain molecular and mechanistic insights into initiation of the RAS-RAF signaling cascade, we developed and used a combination of multiscale simulation and experimental approaches. The influence and impact of the membrane on RAS and RAF proteins is a factor we are just beginning to understand and appreciate in more detail. Molecular simulation is an ideal methodology to further study this complicated relationship between the membrane and associated proteins. Our previous work using Multiscale Machine-learned Modeling Infrastructure investigated different lipid compositions solely around the KRAS4b protein and the interplay between protein behavior and these membrane environments. Multiscale Machine-learned Modeling Infrastructure uses machine learning to couple adjacent simulation scales and has been efficiently scaled across some of the world’s largest high-performance computers. Recently, we have expanded this multiresolution framework to include the all-atom simulation scale and to incorporate the RAF RBDCRD domains. Here, we present the overall analysis results from this new simulation campaign comprising a mixture of RAS and RAF RBDCRD proteins. Approximately 35,000 coarse-grained and 10,000 all-atom molecular dynamics simulations were completed, sampled from a variety of protein/lipid composition configurations that were generated from a micron-scale continuum simulation containing hundreds of copies of the proteins. Our studies suggest that orientations of the RAS-RBDCRD complex on the membrane occupy distinct configurational states, and the spatial patterns of lipid arrangements around these different protein states are unique to each state. The extent and size of lipid “fingerprints” imposed on the membrane by the RAS-RBDCRD protein complex are significantly larger than observed for just the RAS protein on its own. These protein complexes strongly associate, but we do not observe statistically significant preferred protein-protein orientations. These observations indicate that spatial colocalization of RAS-RBDCRD proteins in the same vicinity may be assisted by specific membrane environments, acting to increase the probability of signaling complex formation.

Carpenter, Timothy S. [Lawrence Livermore National↗
DOE OSTI2017

Power demand risk models on milling machines

The measurement of power demand risks in manufacturing power systems will benefit manufacturers and the wider society. If the risks can be characterized using manufacturing parameters, manufacturers can better control risks originating in those parameters, select less power-risky production plans, and reduce utility costs and resource consumption. The measurement of risk can also help manufacturers and power suppliers to protect their power systems from unexpected disturbances. Existing measures of risk, however, do not consider time duration, and thus cannot accurately quantify the risks in manufacturing power systems; the risks of a period of high power demand must be evaluated with the duration of the surge. Therefore, new methods of measuring power demand risks are proposed, adapting measures drawn from the field of finance. With a focus on milling operations, processing power is shown to be a function of processing amount and processing time, and a power demand distribution is directly derived as a joint distribution of and A and T. A bivariate random variable model with copulas is applied to examine the correlation in the joint distribution. Then, based on evaluation of a probability distribution of power demand from and, new risk measures are introduced. Herein illustrative examples are provided to show how the proposed measures can quantify the power demand risks from milling machines, based on manufacturing parameters. Certain manufacturing parameters are found to affect overall power demand risks, including i) raw material type, ii) variability in processing time, and iii) correlation between and. In the examples, these three factors increase power demand risks by up to 108%, 67%, and 1% respectively.

42 ENGINEERING↗
DOE OSTI2019

Temperature gradients for thermophysical and thermochemical property measurements to 3000 °C for an aerodynamically levitated spheroid

This study examines thermal gradients in ceramic oxide spheroids being aerodynamically levitated in a conical nozzle levitator (CNL) system equipped with a CO 2 laser (10.6 μm wavelength). The CNL system is a versatile piece of equipment that can easily be coupled with advanced thermophysical and thermochemical measuring devices, such as diffraction/scattering (X-ray and neutron), nuclear magnetic resonance, and calorimetry, for the analysis of bulk spheroidal solids and liquids. The thermal gradients of a series of single crystal, polycrystalline solids, and liquid spheroids have been measured spatially in the CNL system, by means of a disappearing filament pyrometer (800-3000 °C) and by X-ray diffraction with reference to an internal standard (Pt: 800-1600 °C). The thermal gradient in a levitated sample being heated by a laser from the top can be minimized by: (i) maximizing the sphericity, (ii) maximizing the density, and (iii) minimizing microstructural features. A spheroid with these properties can be manufactured via machining a perfect sphere from a highly dense, chemically and phase pure pellet. These properties promote rotation of the sample about multiple axes in the air stream, enabling homogeneous heating. This homogeneous heating is the dominant factor in reducing thermal gradients in solid state samples. It was found that the thermal gradient in an similar to 3 mm diameter solid sample could be reduced from 1000 °C to 30 °C, by having a perfectly spherical shape that could rotate on multiple axes in a high velocity gas stream (similar to 1500-2000 cm 3 /min). As a result, these findings will allow accurate thermophysical and thermochemical property measurements of solids in situ at high temperatures, using the CNL system.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Reducing the L-H transition power threshold in ITER-similar-shape DIII-D hydrogen plasmas

Recent dedicated DIII-D experiments in low-torque, ITER-similar-shape (ISS) hydrogen plasmas (at a plasma current I p ~ 1.5 MA and ITER-similar edge safety factor q 95 ~ 3.6) show that the L-H transition power threshold P LH can be reduced substantially (~30%) with L-mode helium admixtures n He /n e $\leqslant$ 25%. In the ensuing H-mode, helium ion fractions n He /n H remain below 25%. H-mode normalized pressure and confinement quality are only slightly affected by helium seeding, and Z eff $\leqslant$ 2.15 (including helium and carbon content). The plasmas investigated here are electron-heat dominated, with temperatures T e (0)/T i (0) $\geqslant$ 1 and edge heat flux ratio Q e /Q i (ρ = 0.95) ~ 1.2–1.5. Without mitigation, P LH is higher by a factor of 2–3 in comparison to similar ISS deuterium plasmas. ISS hydrogen plasmas with lower plasma current I p ~ 1 MA (increased edge safety factor q 95 ~ 5.1) exhibit a substantially lower power threshold. This plasma current dependence, also observed previously on ASDEX-U and in JET, is not accounted for by the commonly used 2008 ITPA multi-machine threshold scaling, but could potentially allow H-mode access at marginal heating power during the initial plasma current ramp-up. Attempts to reduce P LH with low-field- and high-field-side hydrogen pellet injection, using 1.7 mm diameter pellets, have not demonstrated a robust threshold reduction, in contrast to successful earlier experiments with larger 2.7 mm pellets. Finally, techniques for reducing P LH are very important for ITER, in particular for accessing H-mode in hydrogen plasmas during the Pre-Fusion Power Operation-1 (PFPO-1) campaign with marginal auxiliary heating power (20–30 MW of ECH).

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

A Self-Evolution Data Fusion Platform for Large-Scale Water Models

Focal Area: Data acquisition and assimilation enabled by machine learning (ML), artificial intelligence (AI), and advanced methods including experimental/network design/optimization, unsupervised learning (deep learning), leveraging advanced hardware (e.g., edge computing). This development is to enable solving the science questions regarding the human and climatic factors that interact with and drive global water scarcity.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Developing a Prototype Methodology to Rank CO2-EOR Wells and Assess Their Reuse Potential for Geologic Carbon Sequestration

This study presents a prototype methodology for evaluating the reuse potential of Class II CO₂-enhanced oil recovery (CO₂-EOR) wells as Class VI wells for geologic carbon sequestration. The approach focuses on assessing wellbore construction materials—casing, cement, tubing, and packers—based on U.S. Environmental Protection Agency (EPA) Class VI well conversion guidelines. Utilizing Python scripts and JSON representations, the methodology automates checks on digitized Texas Railroad Commission (TRRC) data to rank wells based on regulatory and integrity criteria. Key factors include casing integrity, cementing techniques, tubing compatibility, and packer selection. Due to limitations in digitized data, manual verification is required for certain sections. Future enhancements include incorporating non-digitized data via web scraping and machine learning. This research provides a practical framework for well owners and regulators, supporting informed decision-making for sustainable CO₂ storage.

carbon sequestration↗
DOE OSTI2022

ExaTN: Scalable GPU-Accelerated High-Performance Processing of General Tensor Networks at Exascale

We present ExaTN (Exascale Tensor Networks), a scalable GPU-accelerated C++ library which can express and process tensor networks on shared- as well as distributed-memory high-performance computing platforms, including those equipped with GPU accelerators. Specifically, ExaTN provides the ability to build, transform, and numerically evaluate tensor networks with arbitrary graph structures and complexity. It also provides algorithmic primitives for the optimization of tensor factors inside a given tensor network in order to find an extremum of a chosen tensor network functional, which is one of the key numerical procedures in quantum many-body theory and quantum-inspired machine learning. Numerical primitives exposed by ExaTN provide the foundation for composing rather complex tensor network algorithms. We enumerate multiple application domains which can benefit from the capabilities of our library, including condensed matter physics, quantum chemistry, quantum circuit simulations, as well as quantum and classical machine learning, for some of which we provide preliminary demonstrations and performance benchmarks just to emphasize a broad utility of our library.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

INL Senior Project

What did my team set out to accomplish:? Can I put a custom Machine Learning Model on FPGA?? Can I analyze network traffic in real time?? Can a QSFP port be used with an FPGA?? Does a visual representation of the latent space enhance our understanding of network traffic?? What is QSFP QSFP (Quad Small Form-Factor Pluggable)? QSFP supports transfer speeds generally up to 100Gb/s? Runs 4 parallel lines running up to 28 Gb/s? Why the latent space is important to our project ?Latent space is the compressed mapping of data points in a non-linear fashion? Create an understanding of the relationship of data collected? Can represent that relationship of a single network packet in 3 points (X, Y, Z) Project Outline FPGA? Custom Xilinx Petalinux Image for the operating system? Python program to collect packets and run them through the DPU (Data Processing Unit)? The program then sends the information over a socket to a computer? Display Program? Python Program that collects the information sent from the FPGA and display it in a graph

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2017

Design Considerations of a Transverse Flux Machine for Direct-Drive Wind Turbine Applications

This paper presents the design considerations of a double-sided transverse flux machine (TFM) for direct-drive wind turbine applications. The TFM has a modular structure with quasi-U stator cores and ring windings. The rotor is constructed with ferrite magnets in a flux-concentrating arrangement to achieve high air gap flux density. The design considerations for this TFM with respect to initial sizing, pole number selection, key design ratios, and pole shaping are presented in this paper. Pole number selection is critical in the design process of a TFM because it affects both the torque density and power factor under fixed magnetic and changing electrical loading. Several key design ratios are introduced to facilitate the design procedure. The effect of pole shaping on back-emf and inductance is also analyzed. These investigations provide guidance toward the required design of a TFM for direct-drive applications. The analyses are carried out using analytical and three-dimensional finite element analysis. A prototype is under construction for experimental verification.

17 WIND ENERGY↗
DOE OSTI2021

Improving Detection of Disease Re-emergence Using a Web-Based Tool (RED Alert): Design and Case Analysis Study

Background: Currently, the identification of infectious disease re-emergence is performed without describing specific quantitative criteria that can be used to identify re-emergence events consistently. This practice may lead to ineffective mitigation. In addition, identification of factors contributing to local disease re-emergence and assessment of global disease re-emergence require access to data about disease incidence and a large number of factors at the local level for the entire world. This paper presents Re-emerging Disease Alert (RED Alert), a web-based tool designed to help public health officials detect and understand infectious disease re-emergence. Objective: Our objective is to bring together a variety of disease-related data and analytics needed to help public health analysts answer the following 3 primary questions for detecting and understanding disease re-emergence: Is there a potential disease re-emergence at the local (country) level? What are the potential contributing factors for this re-emergence? Is there a potential for global re-emergence? Methods: We collected and cleaned disease-related data (eg, case counts, vaccination rates, and indicators related to disease transmission) from several data sources including the World Health Organization (WHO), Pan American Health Organization (PAHO), World Bank, and Gideon. We combined these data with machine learning and visual analytics into a tool called RED Alert to detect re-emergence for the following 4 diseases: measles, cholera, dengue, and yellow fever. We evaluated the performance of the machine learning models for re-emergence detection and reviewed the output of the tool through a number of case studies. Results: Our supervised learning models were able to identify 82%-90% of the local re-emergence events, although with 18%-31% (except 46% for dengue) false positives. This is consistent with our goal of identifying all possible re-emergences while allowing some false positives. The review of the web-based tool through case studies showed that local re-emergence detection was possible and that the tool provided actionable information about potential factors contributing to the local disease re-emergence and trends in global disease re-emergence. Conclusions: To the best of our knowledge, this is the first tool that focuses specifically on disease re-emergence and addresses the important challenges mentioned above.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

FEDERATED LEARNING ON STOCHASTIC NEURAL NETWORKS

Federated learning is a machine learning paradigm that leverages edge computing on client devices to optimize models while maintaining user privacy by ensuring that local data remain on the device. However, since all data are collected by clients, federated learning is susceptible to latent noise in local datasets. Factors such as limited measurement capabilities or human errors may introduce inaccuracies in client data. To address this challenge, we propose the use of a stochastic neural network as the local model within the federated learning framework. Stochastic neural networks not only facilitate the estimation of the true underlying states of the data but also enable the quantification of latent noise. We refer to our federated learning approach, which incorporates stochastic neural networks as local models, as federated stochastic neural networks. In this work we will present numerical experiments demonstrating the performance and effectiveness of our method, particularly in handling nonindependent and identically distributed data.

97 MATHEMATICS AND COMPUTING↗