Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Generative Model”

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 829 records · Page 46
All sourcesNASA NTRSDOE OSTI
DOE OSTI2015

NREL Evaluates Performance of Hydraulic Hybrid Refuse Vehicles

This highlight describes NREL's evaluation of the in-service performance of 10 next-generation hydraulic hybrid refuse vehicles (HHVs), 8 previous-generation (model year 2013) HHVs, and 8 comparable conventional diesel vehicles operated by Miami-Dade County's Public Works and Waste Management Department in southern Florida. Launched in March 2015, the on-road portion of this 12-month evaluation focuses on collecting and analyzing vehicle performance data - fuel economy, maintenance costs, and drive cycles - from the HHVs and the conventional diesel vehicles. The fuel economy of heavy-duty vehicles, such as refuse trucks, is largely dependent on the load carried and the drive cycles on which they operate. In the right applications, HHVs offer a potential fuel-cost advantage over their conventional counterparts. This advantage is contingent, however, on driving behavior and drive cycles with high kinetic intensity that take advantage of regenerative braking. NREL's evaluation will assess the performance of this technology in commercial operation and help Miami-Dade County determine the ideal routes for maximizing the fuel-saving potential of its HHVs. Based on the field data, NREL will develop a validated vehicle model using the Future Automotive Systems Technology Simulator, also known as FASTSim, to study the impacts of route selection and other vehicle parameters. NREL is also analyzing fueling and maintenance data to support total-cost-of-ownership estimations and forecasts. The study aims to improve understanding of the overall usage and effectiveness of HHVs in refuse operation compared to similar conventional vehicles and to provide unbiased technical information to interested stakeholders.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2016

VALIDATION OF A RELAP5 POINT KINETICS MODEL OF TREAT

Development of advanced nuclear fuels requires testing of fuel specimens under various power-cooling mismatch conditions. The Transient Reactor Test (TREAT) facility at the Idaho National Laboratory will be used to provide transient nuclear heating of fuel specimens. Experiments will be conducted to support assessment of accident tolerant fuel designs being proposed for light water reactor applications. The experiments to be conducted at TREAT will provide data for the assessment of advanced, multi-physics computer codes. However, before the experiments can be performed, calculations are required to demonstrate that they will meet program objectives and that they can be performed safely in TREAT in the event of an unplanned reactivity insertion accident that occurs during the experiment. The safety calculations for the first experiments need to be performed in the next year and therefore will be performed with an existing computer code, RELAP5-3D. A point kinetics model of TREAT was developed for RELAP5-3D. The purpose of the model is to calculate the reactor power during experiments and to support the accident analyses of those experiments. The RELAP5-3D model was validated using data from nine TREAT experiments conducted previously. The RELAP5-3D model generates results that are in reasonable agreement with measured values of maximum core power, energy deposition, and maximum fuel temperature for a wide range of reactivity insertions.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2020

Demand Response Potential from the Bulk Grid Perspective

Gain insight from an exploration of the potential value of demand response (DR) in future power systems. Using hourly DR resource estimates by sub-sector, electrical end-use, and grid service, we incorporate virtual generator models of DR resources into large-scale production cost models, and analyze the grid-level impacts in terms of the ability of DR to reduce system costs and potentially help integrate renewables or reduce system emissions. Additionally, we examine the value of DR, its revenue, and how these change under different system conditions. In this panel session, we will describe our input data and modeling methods, provide an overview of the production cost models used, and summarize key results from analyses in Florida and the western United States. Findings include variations in DR value by end-use and geography, an ability to increase economic PV deployment potential in high penetration systems, and mixed emissions outcomes depending on system composition and fuel prices.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2019

PAVE: An In Situ Framework for Scientific Visualization and Machine Learning Coupling

Machine learning (ML) has emerged as a tool for understanding data at scale. However, this new methodology comes at a cost because ML requires the use of even more HPC resources to generate ML algorithms. In addition to the compute resources required to develop ML algorithms, ML does not sidestep one of the biggest challenges on leading-edge HPC systems: the increasing gap between compute performance and I/O bandwidth. This has led to a strong push towards in situ, processing the data as it is generated, strategies to mitigate the I/O bottleneck. Unfortunately, there are no in situ frameworks dedicated to coupling scientific visualization and ML at scale to develop ML algorithms for scientific visualization. To address the ML and in situ visualization gap, we introduce PAVE. PAVE is an in situ framework which addresses the data management needs between visualisation and machine learning tasks. We demonstrate our framework with a case study that accelerates physically-based light rendering, path-tracing, through the use of a conditional Generative Adversarial neural Network (cGAN). PAVE couples the training over path-traced images resulting in a generative model able to produce scene renderings with accurate light transport and global illumination of a quality comparable to offline approaches in a more efficient manner.

Kim, Mark B.↗
DOE OSTI2020

Urban-scale Energy Modeling: Scaling Beyond Tax Assessor Data

In an attempt to attain building-specific characteristics for urban-scale building energy models, county-specific tax assessors’ data is often an initial data source. This data source can contain valuable information such as year built, area, height, HVAC type, and roof/wall descriptions.We will show examples of 2,000 fields from Hamilton County in Tennessee with examples of many fields which are not relevant to urban-scale building energy modeling, are incorrect compared to other data sources, and highlight some lessons learned working with such a data source.There are currently 3,142 counties in the United States, each with their own data format, field definitions, and data access policy. As urban-scale involves city-scale analysis potentially covering multiple counties and matures toward state- or nation-scale analysis, county-by-county approaches are not scalable. While there are efforts to unify these datasets, there is an increasing proliferation of data and algorithms that can cover wider areas and provide more accurate inputs for urban-scale models. This paper summarizes computer vision of imagery, cartographic layers, building type assessment, and model generation used to achieve scalable detection and analysis of buildings.

New, Joshua↗
DOE OSTI2021

A Modularized Urban Scale Building Energy Modeling Framework Designed with An Open Mind

In recent years, physics-based building energy modeling (BEM) has started being used to evaluate the performance of buildings in the context of connected communities and on an urban scale to study their aggregated energy use, interactions, and impacts on the energy supply infrastructure and environment. The development of urban-scale BEM solutions needs extensive effort. Existing attempts tend to focus on different aspects of BEM on an urban scale, such as collecting as-built building data from different information sources, integrating geometry modeling with geographic information systems (GISs), representing operational and occupancy profiles, automating workflow, processing and visualizing the results, and conducting large-scale simulations. Urban-scale BEM development would benefit from multi-disciplinary research areas and from an open platform to adopt advancements on data sources and tools. For these purposes, this research proposes a modularized bottom-up model creation and simulation framework that is built on the state-of-the-art BEM tools and can accommodate different building stock data. This framework uses a standardized schema to describe building design and operational characteristics, and it can be instantiated from different building survey datasets with heterogeneous structures. The paper demonstrates how thousands of surveyed buildings from the 2012 U.S. Energy Information Administration’s Commercial Buildings Energy Consumption Survey (CBECS) were one-to-one converted to EnergyPlus models through the schema and the model generation process, then simulated with distributed computing, and their results are summarized.

Lei, Xuechen↗
DOE OSTI2022

Uncertainty Quantification in High-Low Dynamic System Coupling using RAVEN and TRANSFORM

This work demonstrates new functionality and applications stemming from the development of high-fidelity to low-fidelity (high-low) coupling for system simulations and to further explore the capabilities of the Risk Analysis Virtual Environment (RAVEN) in the performance of uncertainty quantification in this kind of high-low coupled system models. The work builds from previous work on high-low coupling that utilized COBRA-TF (CTF), the high-fidelity subchannel analysis code, with a low fidelity model built in ORNL’s TRANSFORM, the system analysis code, utilizing the Functional Mock-Up Interface (FMI). Steady-state and transient analysis examples using the high/low coupled models generated from CTF and TRANSFORM/FMI are investigated. The workflows for both steady-state and transient coupled simulations are described. A steady-state parameter sweep and uncertainty analysis of the primary flow rates and reactor power are demonstrated. Likewise, a transient pump trip and power ramp sensitivity studies are also demonstrated. This work elucidates some of the potential benefits and future needs of using RAVEN for high/low system coupling analysis of energy systems. It also shows some of the difficulties that can be encountered in coupling system simulations.

Williams, Wesley↗
DOE OSTI2022

Three Modeling Tools Accelerate the Development of Marine Energy Technologies

Enough power flows through U.S. rivers and oceans to meet up to 60% of the country's electricity needs. But before today's early-stage marine energy technologies can harness some of that power, they need to overcome a few challenges, like a volatile ocean and high development costs. And to do that, technology developers need reliable tools to understand and overcome the challenges they face. That's why researchers at the National Renewable Energy Laboratory (NREL) have developed several open-source, validated, and customizable numerical tools for technology developers. With these tools, the marine energy community can assess their devices' potential power output, predict how their technologies might handle extreme loads (like extreme waves), and learn how changing one component might affect their device's performance. The lab's models generate highly accurate, reliable, robust data that can help marine energy devices leap from the computer screen to the ocean and, eventually, the power grid.

current energy↗
DOE OSTI2022

Learning infinite-horizon average-reward restless multi-action bandits via index awareness

We consider the online restless bandits with average-reward and multiple actions, where the state of each arm evolves according to a Markov decision process (MDP), and the reward of pulling an arm depends on both the current state of the corresponding MDP and the action taken. Since finding the optimal control is typically intractable for restless bandits, existing learning algorithms are often computationally expensive or with a regret bound that is exponential in the number of arms and states. In this paper, we advocate \textit{index-aware reinforcement learning} (RL) solutions to design RL algorithms operating on a much smaller dimensional subspace by exploiting the inherent structure in restless bandits. Specifically, we first propose novel index policies to address dimensionality concerns, which are provably optimal. We then leverage the indices to develop two low-complexity index-aware RL algorithms, namely, (i) GM-R2MAB, which has access to a generative model; and (ii) UC-R2MAB, which learns the model using an upper confidence style online exploitation method. We prove that both algorithms achieve a sub-linear regret that is only polynomial in the number of arms and states. A key differentiator between our algorithms and existing ones stems from the fact that our RL algorithms contain a novel exploitation that leverages our proposed provably optimal index policies for decision-makings.

Xiong, Guojun↗
DOE OSTI2024

Understanding the High Energy Higgs Sector with the CMS Experiment and Artificial Intelligence

This dissertation describes efforts towards understanding the Higgs boson at the highest energies humanly accessible, using the CMS experiment at the Large Hadron Collider and advances in artificial intelligence (AI) and machine learning (ML). We present searches for resonant and nonresonant Higgs-boson (H) pair production in the all-hadronic two beauty-quark and two vector boson (V) final state, using a novel strategy to measure the quartic HHVV coupling and search for new Higgs-like bosons. By targeting highly Lorentz-boosted Higgs pairs, we probe effects of potential new physics in the high energy Higgs sector, which could hold answers to fundamental mysteries of nature such as baryon asymmetry. To enable these and future searches, we introduce as well significant developments in AI/ML, including in the identification of boosted H$\rightarrow$VV decays with deep transformer networks and advances in AI-accelerated fast simulations of the CMS detector. The latter notably includes the development of the first, highly performant generative models for point-cloud data in high energy physics, which have the potential to improve CMS' computational efficiency by up to three orders of magnitude. We also highlight novel solutions to the important and challenging problems of calibrating and validating these ML techniques. Finally, we present new approaches to search for new physics in a model-agnostic manner, using physics-informed ML methods equivariant to Lorentz transformations. The quartic HHVV coupling is observed (expected) to be constrained to $[-0.04, 2.05]$ ($[0.05, 1.98]$) at the 95% confidence level relative to the standard model prediction, representing the second-most sensitive measurement of this coupling by CMS to date. Exclusion limits on the production cross section of new heavy resonances decaying to two Higgs-like bosons are expected to be as low as 0.3 fb for high resonance masses.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

CAD-VAE: Leveraging Correlation-Aware Latents for Comprehensive Fair Disentanglement

While deep generative models have significantly advanced representation learning, they may inherit or amplify biases and fairness issues by encoding sensitive attributes alongside predictive features. Enforcing strict independence in disentanglement is often unrealistic when target and sensitive factors are naturally correlated. To address this challenge, we propose CAD-VAE(Correlation-Aware Disentangled VAE), which introduces a correlated latent code to capture the information shared between the target and sensitive attributes. Given this correlated latent, our method effectively separates over-lapping factors without extra domain knowledge by directly minimizing the conditional mutual information between target and sensitive codes. A relevance-driven optimization strategy refines the correlated code by efficiently capturing essential correlated features and eliminating redundancy. Extensive experiments on benchmark datasets demonstrate that CAD-VAE produces fairer representations, realistic counterfactuals, and improved fairness-aware image editing.

Ma, Chenrui [University of California Irvine]↗
DOE OSTI2022

PolyODENet

Kinetics of a reaction network that follows mass-action rate laws can be described with a system of ordinary differential equations (ODEs) with polynomial right-hand side. However, it is challenging to derive such kinetic differential equations from transient kinetic data without knowing the reaction network, especially when the data are incomplete due to experimental limitations. We introduce a program, PolyODENet, toward this goal. Based on the machine-learning method Neural ODE, PolyODENet defines a generative model and predicts concentrations at arbitrary time. As such, it is possible to include unmeasurable intermediate species in the kinetic equations. Importantly, we have implemented various measures to apply physical constraints and chemical knowledge in the training to regularize the solution space.

Wu, Qin [Brookhaven National Lab. (BNL), Upton, NY↗
DOE OSTI2010

Standard model with four generations: Selected implications for rare B and K decays

We extend our recent work and study implications of the standard model with four generations (SM4) for rare B and K decays. We again take seriously the several 2-3{sigma} anomalies seen in B, B{sub s} decays and interpret them in the context of this simple extension of the SM. SM4 is also of course of considerable interest for its potential relevance to dynamical electroweak symmetry breaking and to baryogenesis. Using experimental information from processes such as B{yields}X{sub s}{gamma}, B{sub d} and B{sub s} mixings, indirect CP-violation from K{sub L}{yields}{pi}{pi} etc. along with oblique corrections, we constrain the relevant parameter space of the SM4, and find m{sub t}{sup '} of about 400-600 GeV with a mixing angle |V{sub t}{sup '}{sub b}{sup *}V{sub t}{sup '}{sub s}| in the range of about 0.05-1.4x10{sup -2} and with an appreciable CP-odd associated phase, are favored by the current data. Given the unique role of the CP asymmetry in B{sub s}{yields}{psi}{phi} due to its gold-plated nature, correlation of that with many other interesting observables, including the semileptonic asymmetry (A{sub SL}) are studied in SM4. We also identify several processes, such as B{yields}X{sub s{nu}{nu}}, K{sub L{yields}{pi}}{sup 0{nu}{nu}} etc., that are significantly different in SM4 from the SM. Experimentally the very distinctive process B{sub s}{yields}{mu}{sup +}{mu}{sup -} is also discussed; the branching ratio can be larger or smaller than in SM, (3.2{yields}4.2)x10{sup -9}, by a factor of O(3).

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

High-Level Model Articulation with BuildingSync and OpenStudio

The use of the BuildingSync schema for describing the contents of buildings is becoming more common with its recent integration into the Audit Template tool as well as ASHRAE’s Building Energy Quotient (bEQ) web portal. Although BuildingSync was initially created to store and transfer data related to building energy audits (as defined by ASHRAE Standard 211), it has since been expanded to store the data needed to articulate fully defined physics-based building energy models. BuildingSync combined with abstracted high-level input methods defined in OpenStudio’s Standards project and the newly developed BuildingSync gem allows BuildingSync eXtensible Markup Language (XML) (Bray, Paoli, Sperberg-McQueen, Maler, Eve (Sun Microsystems, & Francois, 2008) document to be converted into OpenStudio models. Automatic generation of a building energy model from data collected during an audit 1) eliminates the need to separately generate an energy model, 2) improves consistency between model generation, and 3) simplifies evaluation of various energy efficiency measures. This manuscript will discuss this open-source project, including: development of the validation infrastructure necessary to provide formalized expectations of informational requirements for BuildingSync documents; development of the BuildingSync gem for translation of BuildingSync documents to OpenStudio models as well as example implementations; and finally, the manuscript will elaborate on the advantages and disadvantages of using high-level models generated from BuildingSync as surrogates to detailed models.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2025

Optimizing flow condensation models for next-generation refrigerants in axial micro-fin aluminum tubes

To support the transition to next-generation refrigerants, accurate modelling of heat transfer and pressure drop is essential for designing efficient heat exchangers. Current models, largely based on traditional refrigerants and unexpanded micro-fin tubes, may not reliably predict performance for new refrigerants and expanded micro-fin geometries. This study evaluates four condensation models using experimental data for six A2L refrigerants: R-32, R-454B, R-454C, R-455A, R-1234yf, and R-1234ze(E). For heat transfer models, the Han and Lee (2005) model initially yields the best accuracy (mean absolute Deviation, MAD = 22.1%). To further improve predictions, a correction factor reduces the Cavallini et al. (2009) model’s MAD from 68.2% to 15.4%, while optimization of the Kedzierski and Goncalves (1997) model achieves a MAD of 13.1%. For pressure drop, the Cavallini et al. (1997) model proves most accurate (MAD = 6.4%), with the simpler Haraguchi et al. (1993) model also effective (MAD = 9.4%). Keywords: Flow condensation models, heat transfer coefficient, frictional pressure drop, next-generation refrigerants, aluminum micro-fin tubes

Hu, Yifeng [ORNL] (ORCID:0000000242875185)↗
DOE OSTI2019

Including operational aspects in the planning of power systems with large amounts of variable generation: A review of modeling approaches

In the past, power system planning was based on meeting the load duration curve at minimum cost. The increasing share of variable generation (VG) makes operational constraints more important in the planning problem, and there is more and more interest in considering aspects such as sufficient ramping capability, sufficient reserve procurement, power system stability, storage behavior, and the integration of other energy sectors often through demand response assets. In VG integration studies, several methods have been applied to combine the planning and operational timescales. We present a four-level categorization for the modeling methods, in order of increasing complexity: (1a) investment model only, (1b) operational model only, (2) unidirectionally soft-linked investment and operational models, (3a) bidirectionally soft-linked investment and operational models, (3b) operational model with an investment update algorithm, and (4) co-optimization of investments and operation. The review shows that using a low temporal resolution or only few representative days will not suffice in order to determine the optimal generation portfolio. In addition, considering operational effects proves to be important in order to get a more optimal generation portfolio and more realistic estimations of system costs. However, operational details appear to be less significant than the temporal representation. Furthermore, the benefits and impacts of more advanced modeling techniques on the resulting generation capacity mix significantly depend on the system properties. Thus, the choice of the model should depend on the purpose of the study as well as on system characteristics.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2017

A New Generation of Standard Solar Models

We compute a new generation of standard solar models (SSMs) that includes recent updates on some important nuclear reaction rates and a more consistent treatment of the equation of state. Models also include a novel and flexible treatment of opacity uncertainties based on opacity kernels, required in light of recent theoretical and experimental works on radiative opacity. Two large sets of SSMs, each based on a different canonical set of solar abundances with high and low metallicity (Z), are computed to determine model uncertainties and correlations among different observables. We present detailed comparisons of high- and low-Z models against different ensembles of solar observables, including solar neutrinos, surface helium abundance, depth of the convective envelope, and sound speed profile. A global comparison, including all observables, yields a p-value of 2.7σ for the high-Z model and 4.7σ for the low-Z one. When the sound speed differences in the narrow region of 0.65<r/R{sub ⊙}<0.70 are excluded from the analysis, results are 0.9σ and 3.0σ for high- and low-Z models respectively. These results show that high-Z models agree well with solar data but have a systematic problem right below the bottom of the convective envelope linked to steepness of molecular weight and temperature gradients, and that low-Z models lead to a much more general disagreement with solar data. We also show that, while simple parametrizations of opacity uncertainties can strongly alleviate the solar abundance problem, they are insufficient to substantially improve the agreement of SSMs with helioseismic data beyond that obtained for high-Z models due to the intrinsic correlations of theoretical predictions.

79 ASTRONOMY AND ASTROPHYSICS↗