SEARCH · Engineering Papers

Results for “Real numbers”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 757 records · Page 42

All sources NASA NTRS DOE OSTI

DOE OSTI2023

EI_MS_ML

The unambiguous identification of compounds from their electron ionization mass (EI-MS) spectra remains a significant unsolved problem in the field of metabolomics and analytical chemistry as a whole. Typically EI-MS spectra are compared using various mathematical operations that convert the spectral similarity or differences into a distance-like metric that roughly approximates the similarity of any two spectra. A commonly used metric for this is the cosine similarity metric which has values close to one for very similar spectra and a value of zero for very dissimilar spectra; however, no metric is perfect. Due to the prevalence of structurally-similar compounds such as isomers and the prevalence of certain fragmentation patterns across structurally-dissimilar compounds, the unambiguous assignment of EI-MS spectra compounds remains difficult. Frequently, querying an observed EI-MS spectrum against a large database such as the NIST17 library yields multiple possible assignments requiring the end user to distinguish between multiple high scoring hits, or multiple low scoring hits while keeping in mind that the correct hit may not be in the database at all. Although techniques such as orthogonal information from techniques such as chromatography can greatly aid in unambiguous assignment, this also requires more complicated experimental designs and access to more complicated analytical instrumentation. Substructures can be trivially detected and represented as strings using a previously published technique called node coloring from a known chemical structure. However, for experimentally-derived EI-MS spectra this information must be derived from the spectra itself (i.e., because we do not know what compound it represents). To achieve this, the software uses techniques from the field of machine learning and a large training dataset of EI-MS spectra corresponding to known structures annotated with substructure strings, to build models that can predict the presence of a given chemical substructure from an EI-MS spectrum directly.If these predictions are of high-quality (i.e., are unlikely to be false positives), the presence of one or more predicted substructures can be used to constrain the number of possible hits for a query spectrum. Mathematically, this restriction could be expressed in many forms, but the most straight-forward implementation is to weight the cosine similarity of a query spectrum and a plausible database match with a Tanimoto-like coefficient based on the ratio of the number of substructures predicted to the number of substructures present in the potential database hit. Determining which combination of models best reduces assignment ambiguity will be achieved using a combination of manual curation and optimization techniques such as genetic algorithms. This software will perform all the steps necessary to construct said models from a training dataset and evaluate them using a holdout dataset. Various statistical analyses can be performed to determine if this approach does decrease assignment ambiguity. For example, if this approach works, on average, the rank-order of the correct assignment for the holdout set of EI-MS spectra should decrease and the weighted cosine similarities for most of the possible matches in the database should be better than the unweighted cosine similarities. Furthermore, this same pipeline can be used on real experimental data to generate less ambiguous assignments.

Mitchell, Joshua↗

DOE OSTI2024

Challenges of COVID-19 Case Forecasting in the US, 2020–2021

During the COVID-19 pandemic, forecasting COVID-19 trends to support planning and response was a priority for scientists and decision makers alike. In the United States, COVID-19 forecasting was coordinated by a large group of universities, companies, and government entities led by the Centers for Disease Control and Prevention and the US COVID-19 Forecast Hub ( https://covid19forecasthub.org ). We evaluated approximately 9.7 million forecasts of weekly state-level COVID-19 cases for predictions 1–4 weeks into the future submitted by 24 teams from August 2020 to December 2021. We assessed coverage of central prediction intervals and weighted interval scores (WIS), adjusting for missing forecasts relative to a baseline forecast, and used a Gaussian generalized estimating equation (GEE) model to evaluate differences in skill across epidemic phases that were defined by the effective reproduction number. Overall, we found high variation in skill across individual models, with ensemble-based forecasts outperforming other approaches. Forecast skill relative to the baseline was generally higher for larger jurisdictions (e.g., states compared to counties). Over time, forecasts generally performed worst in periods of rapid changes in reported cases (either in increasing or decreasing epidemic phases) with 95% prediction interval coverage dropping below 50% during the growth phases of the winter 2020, Delta, and Omicron waves. Ideally, case forecasts could serve as a leading indicator of changes in transmission dynamics. However, while most COVID-19 case forecasts outperformed a naïve baseline model, even the most accurate case forecasts were unreliable in key phases. Further research could improve forecasts of leading indicators, like COVID-19 cases, by leveraging additional real-time data, addressing performance across phases, improving the characterization of forecast confidence, and ensuring that forecasts were coherent across spatial scales. In the meantime, it is critical for forecast users to appreciate current limitations and use a broad set of indicators to inform pandemic-related decision making.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2024

How well are hazards associated with derechos reproduced in regional climate simulations?

Abstract. A 15-member ensemble of convection-permitting regional simulations of the fast-moving and destructive derecho of 29–30 June 2012 that impacted the northeastern urban corridor of the USA is presented. This event generated 1100 reports of damaging winds, generated significant wind gusts over an extensive area of up to 500 000 km2, caused several fatalities, and resulted in widespread loss of electrical power. Extreme events such as this are increasingly being used within pseudo-global-warming experiments to examine the sensitivity of historical, societally important events to global climate non-stationarity and how they may evolve as a result of changing thermodynamic and dynamic contexts. As such it is important to examine the fidelity with which such events are described in hindcast experiments. The regional simulations presented herein are performed using the Weather Research and Forecasting (WRF) model. The resulting ensemble is used to explore simulation fidelity relative to observations for wind gust magnitudes, spatial scales of convection (as is manifest in high composite reflectivity, cREF), and both rainfall and hail production as a function of model configuration (microphysics parameterization, lateral boundary conditions (LBCs), start date, use of nudging, compiler choice, damping, and number of vertical levels). We also examine the degree to which each ensemble member differs with respect to key mesoscale drivers of convective systems (e.g., convective available potential energy and vertical wind shear) and critical manifestations of deep convection, e.g., vertical velocities, cold-pool generation, and how those properties relate to the correct characterization of the associated atmospheric hazards (wind gusts and hail). Use of a double-moment, seven-class scheme with number concentrations for all species (including hail and graupel) results in the greatest fidelity of model-simulated wind gusts and convective structure to the observations of this event. All ensemble members, however, fail to capture the intensity of the event in terms of the spatial extent of convection and the production of high near-surface wind gusts. We further show very high sensitivity to the LBCs employed and specifically that simulation fidelity is higher for simulations nested within ERA-Interim compared to ERA5. Excess convective available potential energy (CAPE) in all ensemble members after the derecho passage leads to excess production of convective cells, wind gusts, cREF > 40 dBZ, and precipitation during a frontal passage on the subsequent day. This event proved very challenging to forecast in real time and to reproduce in the 15-member hindcast simulation ensemble presented here. Future work could examine if simulations with other initial and lateral boundary conditions can achieve greater fidelity.

Shepherd, Tristan (ORCID:0000000186276419)↗

DOE OSTI2020

Neural network based prediction of no-wall β N limits due to ideal external kink instabilities

Multi-layer perceptron based neural networks (NNs) are trained to predict the Troyon no-wall beta limits due to the onset of low-n (n = 1 – 3, n is the toroidal mode number) ideal external kink instabilities in tokamak plasmas. It is revealed that a well-trained NN can predict the n = 1 no-wall beta limit within 10% relative error. The NN performance is somewhat worse for the n = 2 and 3 no-wall beta limits, but still a relative error within 20% is achievable. The trained NNs well reproduce the known dependences of the no-wall beta limits on the plasma pressure (pressure peaking factor) and current (plasma internal inductance) profiles. Other scalings are further easily established with NNs, for parametric dependences such as on the aspect ratio, the elongation and triangularity of the plasma boundary shape. These NN-based Troyon beta limit predictors can be incorporated into integrated modeling platforms, or directly implemented as a real time stability estimator for the purpose of disruption avoidance or mitigation during experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2022

A Review of Modeling Approaches for Understanding and Monitoring the Environmental Effects of Marine Renewable Energy

Understanding the environmental effects of marine energy (ME) devices is fundamental for their sustainable development and efficient regulation. However, measuring effects is difficult given the limited number of operational devices currently deployed. Numerical modeling is a powerful tool for estimating environmental effects and quantifying risks. It is most effective when informed by empirical data and coordinated with the development and implementation of monitoring protocols. We reviewed modeling techniques and information needs for six environmental stressor–receptor interactions related to ME: changes in oceanographic systems, underwater noise, electromagnetic fields (EMFs), changes in habitat, collision risk, and displacement of marine animals. This review considers the effects of tidal, wave, and ocean current energy converters. We summarized the availability and maturity of models for each stressor–receptor interaction and provide examples involving ME devices when available and analogous examples otherwise. Models for oceanographic systems and underwater noise were widely available and sometimes applied to ME, but need validation in real-world settings. Many methods are available for modeling habitat change and displacement of marine animals, but few examples related to ME exist. Models of collision risk and species response to EMFs are still in stages of theory development and need more observational data, particularly about species behavior near devices, to be effective. We conclude by synthesizing model status, commonalities between models, and overlapping monitoring needs that can be exploited to develop a coordinated and efficient set of protocols for predicting and monitoring the environmental effects of ME.

16 TIDAL AND WAVE POWER↗

DOE OSTI2025

Assessing the Potential Impact of Fugitive Methane Emissions on Offshore Platform Safety

One of the biggest risks to safety on offshore platform safety is the ignition of high-pressure natural gas streams. Currently, the size and number of fugitive emissions on offshore platforms is unknown and methods used to detect fugitives have significant shortcomings. To investigate the frequency, size, and potential impact of fugitives, a data collection exercise was conducted using incidents reported, leak survey data, and independent measurements. The size and number of fugitives on offshore facilities were simulated to investigate likely areas of safety concern. Incident reports indicate in 2021 there were 113 reports of gas leaks on 1119 offshore facilities, suggesting 0.02 fugitives per Type 1 facility (older, shallow-water platforms) and 0.31 fugitives per Type 2 facility (larger deeper-water facilities). Leak survey data report 12 fugitives per Type 1 facility (average emission 0.6 kg CH 4 h −1 leak −1 ) and 15 fugitives per Type 2 facility (average emission 1.5 kg CH 4 h −1 leak −1 ). Reconciliation of direct measurements with a bottom-up model suggests that the number of fugitive emissions generated from the leak report data is an underestimate for Type 1 platforms (44 fugitives facility −1 ; average emission 0.6 kg CH 4 h −1 leak −1 ) and in general agreement for the Type 2 platforms (15 fugitives facility −1 ; average emission 1.5 kg CH 4 h −1 leak −1 ). Analysis of the fugitive emission rates on an offshore platform suggests that gas will not collect to explosive concentration if any air movement is present (>0.36 mph); however, large volumes of air (~600 m 3 ) near representative leaks on the working deck could become explosive in hour-long zero-wind conditions. We suggest that wearable technology could be employed to indicate gas build up, safety regulations amended to consider low-wind conditions and real-world experiments are conducted to test assumptions of air mixing on the working deck.

explosion↗

DOE OSTI2022

A hybrid architecture for volt-var control in active distribution grids

Modern active distribution grids are characterized by the increasing penetration of distributed energy resources (DERs). The proper coordination and scheduling of a large numbers of these small-scale and spatially distributed DERs is necessary, and warrants the use of novel distributed approaches. In this paper, we propose a hybrid volt-var control architecture for the distribution grid, which leverages existing centralized and local approaches to planning, decision making, and control, and augments it with distributed optimization and distributed control for DER management. First, we propose a convex model to describe the power physics of distribution grids of meshed topology and unbalanced structure, based on current injection and McCormick Envelopes. Second, we employ the distributed proximal atomic coordination (PAC) algorithm to coordinate DERs to provide voltage support. We implement volt-var optimization by optimally coordinating DERs including PV smart inverters and demand response. We present results using the IEEE-34 bus network, using real data from a distribution feeder in Hawaii, to model load and PV generation. Different levels of DER penetration and objective functions are simulated. Finally, our results show the need for the coordination of DERs to improve voltage profiles, even in networks with existing voltage control devices. Further, we show the need for flexible reactive power capabilities to achieve desired grid performance.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2024

PlantSegNet: 3D point cloud instance segmentation of nearby plant organs with identical semantics

In this study, we introduce PlantSegNet, a novel neural network model for instance segmentation of nearby objects with similar geometric structures. Our work addresses the challenges of instance segmentation of plant point clouds, including the difficulty of annotating and labeling point clouds, the loss of local structural information in neural network components, and the generation of large numbers of incorrect small clusters due to poor choices of the loss function. One of the key contributions of our approach is a digital twin of sorghum, i.e., a procedural sorghum model, which was used to generate point clouds of sorghum fields. This allowed us to create a large-scale, annotated, synthetic dataset of sorghum plants that we used to train our PlantSegNet model. We demonstrated the effectiveness of our method in segmenting instances of sorghum leaves grown in outdoor field settings. To the best of our knowledge, this is the first study to address this specific instance segmentation problem for plants grown in such a setting. We compared our proposed method with other state-of-the-art methods for indoor settings, including SGPN and TreePartNet, on both synthetic and real data. Furthermore, our results show that PlantSegNet outperforms these methods regarding accuracy, robustness, and efficiency.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Production of alternate realizations of DESI fiber assignment for unbiased clustering measurement in data and simulations

A critical requirement of spectroscopic large scale structure analyses is correcting for selection of which galaxies to observe from an isotropic target list. This selection is often limited by the hardware used to perform the survey which will impose angular constraints of simultaneously observable targets, requiring multiple passes to observe all of them. In SDSS this manifested solely as the collision of physical fibers and plugs placed in plates. In DESI, there is the additional constraint of the robotic positioner which controls each fiber being limited to a finite patrol radius. A number of approximate methods have previously been proposed to correct the galaxy clustering statistics for these effects, but these generally fail on small scales. To accurately correct the clustering we need to upweight pairs of galaxies based on the inverse probability that those pairs would be observed (Bianchi & Percival 2017). This paper details an implementation of that method to correct the Dark Energy Spectroscopic Instrument (DESI) survey for incompleteness. To calculate the required probabilities, we need a set of alternate realizations of DESI where we vary the relative priority of otherwise identical targets. These realizations take the form of alternate Merged Target Ledgers (AMTL), the files that link DESI observations and targets. We present the method used to generate these alternate realizations and how they are tracked forward in time using the real observational record and hardware status, propagating the survey as though the alternate orderings had been adopted. We detail the first applications of this method to the DESI One-Percent Survey (SV3) and the DESI year 1 data. We include evaluations of the pipeline outputs, estimation of survey completeness from this and other methods, and validation of the method using mock galaxy catalogs.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2021

Design Considerations for GPU-based Mixed Integer Programming on Parallel Computing Platforms

Mixed Integer Programming (MIP) is a powerful abstraction in combinatorial optimization that finds real-life application across many significant sectors. The recent proliferation of graphical processing unit (GPU)-based accelerated computing architectures in large-scale parallel computing or supercomputing presents new opportunities as well as challenges in the advancement of MIP solver technology to effectively use the new accelerated computing platforms and scale to large parallel systems. Here, we recount the conventional processor-based strategies and focus on configurations where the most promising intersection lies between parallel MIP solver approaches and the specific strengths of accelerated parallel platforms. We note that the best potential lies in solving problems whose individual matrix sizes (of the linear program relaxation) fit entirely within one accelerator's memory and whose branch-and-bound (or branch-and-cut) trees cannot be fully contained within a small number of computational nodes. Additionally, we identify ideal features of computational linear algebra support on GPU accelerators that would help advance this direction of scalable parallel solution of MIP problems on GPU-based accelerated computing architectures.

Perumalla, Kalyan↗

DOE OSTI2025

Evaluation of the Energy, Hygrothermal, and Thermal Capacity Performance of Cross-Laminated Timber

Cross-laminated timber (CLT) construction is gaining momentum in the US because it offers multiple advantages over traditional construction methods. Benefits that have received the most attention focus on constructability, the environment, and protection (e.g., blast resistance), although CLT construction is likely to offer other benefits, as well. Still, these have not been studied at length because such evaluations are costly, requiring long-term assessments in an actual building and specialized technical knowledge. Among the possible benefits, CLT construction likely provides a higher-performing building envelope. Using CLT panels to enclose a building means fewer joints in the opaque envelope than what is required in traditional stick-framed construction. Fewer joints mean fewer locations where the air- and water-resistive barrier (WRB) could be compromised; thus, a CLT building enclosure may require less maintenance and have a longer lifespan than a traditionally built structure because of fewer air and water leaks. In addition, CLT’s thermal mass moderates indoor temperatures, allowing the heating, ventilation, and air conditioning (HVAC) system to operate more efficiently during peak hours, reducing operational energy consumption throughout the lifetime of the CLT building (Salonvaara et al., 2022). Furthermore, more stable indoor temperatures can increase occupant comfort. The CLT’s thermal mass can also reduce energy costs by adjusting to utility time-of-use pricing without affecting occupant comfort. The ability of CLT buildings to bridge periods without HVAC operation prepares them for future grid interaction and provides a certain level of resilience against power outages. Researchers have attempted to quantify these benefits; however, their work is based on simplified simulations with numerous assumptions. To correctly understand the benefits, an actual building must be monitored. Therefore, information needs to be gathered on indoor and outdoor temperatures, HVAC energy consumption, thermostat setpoints, temperatures, and thermal transport in CLT components to comprehend how these parameters are affected by the CLT’s thermal mass. These data are needed to reduce the number of assumptions and calibrate simulation models to optimize HVAC controls to minimize overall energy consumption, reduce energy use and higher fees during peak demand, and maintain occupant comfort. Additionally, the calibrated simulation model allows the optimization exercise to be repeated in various US climates. Potential benefits can be tailored to buildings in various locations, and decisions can be made on where CLT construction could be most advantageous. Furthermore, monitoring and simulation results are needed to evaluate the durability of the CLT structures in different climates. This project’s researchers gathered information to help understand and quantify the benefits of CLT buildings concerning operational energy, moderated indoor temperatures, and comfort; the dynamic operation to provide grid services; and resilience in times of power outage. Through the corroboration of simulation models with real-world measurements, this study paves the way for extrapolating findings to other climatic zones and building typologies, thereby broadening the understanding of CLT’s multifaceted benefits and reinforcing its position as a material of choice in sustainable construction.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2023

Seasonal variations in composition and sources of atmospheric ultrafine particles in urban Beijing based on near-continuous measurements

Abstract. Understanding the composition and sources of atmospheric ultrafine particles (UFPs) is essential in evaluating their exposure risks. It requires long-term measurements with high time resolution, which are scarce to date. We performed near-continuous measurements of UFP composition during four seasons in urban Beijing using a thermal desorption chemical ionization mass spectrometer, accompanied by real-time size distribution measurements. We found that UFPs in urban Beijing are dominated by organic components, varying seasonally from 68 % to 81 %. CHO organics (i.e., molecules containing carbon, hydrogen, and oxygen) are the most abundant in summer, while sulfur-containing organics, some nitrogen-containing organics, nitrate, and chloride are the most abundant in winter. With the increase of particle diameter, the contribution of CHO organics decreases, while that of sulfur-containing and nitrogen-containing organics, nitrate, and chloride increases. Source apportionment analysis of the UFP organics indicates contributions from cooking and vehicle sources, photooxidation sources enriched in CHO organics, and aqueous/heterogeneous sources enriched in nitrogen- and sulfur-containing organics. The increased contributions of cooking, vehicle, and photooxidation components are usually accompanied by simultaneous increases in UFP number concentrations related to cooking emission, vehicle emission, and new particle formation, respectively, while the increased contribution of the aqueous/heterogeneous composition is usually accompanied by the growth of UFP mode diameters. The highest UFP number concentrations in winter are due to the strongest new particle formation, the strongest local primary particle number emissions, and the slowest condensational growth of UFPs to larger sizes. This study provides a comprehensive understanding of urban UFP composition and sources and offers valuable datasets for the evaluation of UFP exposure risks.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2020

Model-Free State Estimation Using Low-Rank Canonical Polyadic Decomposition

As electric grids experience high penetration levels of renewable generation, fundamental changes are required to address real-time situational awareness. Here, we utilize unique traits of tensors to devise a model-free situational awareness and energy forecasting framework for distribution networks. This work formulates the state of the network at multiple time instants as a three-way tensor; hence, recovering full state information of the network is tantamount to estimating all the values of the tensor. Given measurements received from µphasor measurement units and/or smart meters, the recovery of unobserved quantities is carried out using the low-rank canonical polyadic decomposition of the state tensor—that is, the state estimation task is posed as a tensor imputation problem utilizing observed patterns in measured (sampled) quantities. Two structured sampling schemes are considered, namely, asynchronous slab and fiber sampling. For both schemes, we present sufficient conditions on the number of sampled slabs and fibers that guarantee identifiability of the factors of the state tensor. Numerical results demonstrate the ability of the proposed framework to achieve high estimation accuracy in multiple sampling scenarios.

42 ENGINEERING↗

DOE OSTI2022

FY 2021 Isolated Grids and Grid-Connected Turbine Reference Systems; Microgrids, Infrastructure Resilience, and Advanced Controls Launchpad (MIRACL)

For individuals, businesses, and communities focused on building resilient electrical grid infrastructure, wind energy can provide an affordable, accessible, and compatible distributed energy resource option that also enhances the capabilities of local grid operations. The Microgrids, Infrastructure Resilience, and Advanced Controls Launchpad (MIRACL) project is a multi-year distributed wind research effort, driven through a partnership between four Department of Energy National Laboratories and industry to develop and improve the planning, design, and operation of wind-centered microgrids to complement solar, energy storage, and other distributed energy resources for grid-tied and isolated operation (U.S. Department of Energy, 2021). This report documents the application of methods developed through the initial three years of the MIRACL project to two real-world distributed wind reference systems. Specifically, the methods demonstrated in this report include 1) a market valuation framework to comprehensively value the services distributed wind can provide and 2) a resilience framework that enables stakeholders to characterize distribution system resilience and compare grid investment decisions from a resilience perspective. Additional methods mentioned in this report include distributed hybrid system design methods for grid resilience, advanced control co-simulation platforms, and power hardware-in-the-loop (PHIL) models. Preliminary results from these additional methods are presented in this report and will be demonstrated and/or applied to the reference systems in the coming year. The purpose of applying these methods to reference systems is to drive technology transfer of the theories, methodologies, and technologies developed under the MIRACL project and increase the number of referenceable case studies available to stakeholders interested in additional value-added capabilities of wind systems beyond bulk energy supply (i.e. kilowatt-hours).

17 WIND ENERGY↗

DOE OSTI2023

Constraints on Populations of Neutrino Sources from Searches in the Directions of IceCube Neutrino Alerts

Beginning in 2016, the IceCube Neutrino Observatory has sent out alerts in real time containing the information of high-energy ($E$ ≳ 100 TeV) neutrino candidate events with moderate to high (≳30%) probability of astrophysical origin. In this work, we use a recent catalog of such alert events, which, in addition to events announced in real time, includes events that were identified retroactively and covers the time period of 2011–2020. We also search for additional, lower-energy neutrinos from the arrival directions of these IceCube alerts. We show how performing such an analysis can constrain the contribution of rare populations of cosmic neutrino sources to the diffuse astrophysical neutrino flux. After searching for neutrino emission coincident with these alert events on various timescales, we find no significant evidence of either minute-scale or day-scale transient neutrino emission or of steady neutrino emission in the direction of these alert events. This study also shows how numerous a population of neutrino sources has to be to account for the complete astrophysical neutrino flux. Assuming that sources have the same luminosity, an $E$ -2.5 neutrino spectrum, and number densities that follow star formation rates, the population of sources has to be more numerous than 7 × 10 -9 Mpc -3 . This number changes to 3 × 10 -7 Mpc -3 if number densities instead have no cosmic evolution.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2023

HOLISMOKES IX. Neural network inference of strong-lens parameters and uncertainties from ground-based images

Modeling of strong gravitational lenses is a necessity for further applications in astrophysics and cosmology. With the large number of detections in current and upcoming surveys, such as the Rubin Legacy Survey of Space and Time (LSST), it is pertinent to investigate automated and fast analysis techniques beyond the traditional and time-consuming Markov chain Monte Carlo sampling methods. Building upon our (simple) convolutional neural network (CNN), we present here another CNN, specifically a residual neural network (ResNet), that predicts the five mass parameters of a singular isothermal ellipsoid (SIE) profile (lens center x and y, ellipticity e x and e y , Einstein radius θ E ) and the external shear (γ ext, 1 , γ ext, 2 ) from ground-based imaging data. In contrast to our previous CNN, this ResNet further predicts the 1σ uncertainty for each parameter. To train our network, we use our improved pipeline to simulate lens images using real images of galaxies from the Hyper Suprime-Cam Survey (HSC) and from the Hubble Ultra Deep Field as lens galaxies and background sources, respectively. We find very good recoveries overall for the SIE parameters, especially for the lens center in comparison to our previous CNN, while significant differences remain in predicting the external shear. From our multiple tests, it appears that most likely the low ground-based image resolution is the limiting factor in predicting the external shear. Given the run time of milli-seconds per system, our network is perfectly suited to quickly predict the next appearing image and time delays of lensed transients. Therefore, we use the network-predicted mass model to estimate these quantities and compare to those values obtained from our simulations. Unfortunately, the achieved precision allows only a first-order estimate of time delays on real lens systems and requires further refinement through follow-up modeling. Nonetheless, our ResNet is able to predict the SIE and shear parameter values in fractions of a second on a single CPU, meaning that we are able to efficiently process the huge amount of galaxy-scale lenses expected in the near future.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

A Hardware-in-the-Loop Experimental Testbed Using Air Conditioners for Grid Balancing

Driven by the need to offset the variability of renewable generation on the grid, development of load control is a highly active field of research. However, practical use of residential loads for grid balancing remains rare, in part due to the cost of communicating with large numbers of small loads and also the limited experimentation done so far to demonstrate reliable operation. To establish a basis for the safe and reliable use of fleets of compressor loads as distributed energy resources, we constructed an experimental testbed in a laboratory, so that load coordination schemes could be tested at extreme conditions. Here, this experimental testbed was used to tune a simulation testbed to which it was then linked, thereby augmenting the effective size of the fleet. Modeling of the system was done both to demonstrate the experimental testbed's behavior and also to understand how to tune the behavior of each load. Implementing this testbed has enabled rapid turnaround of experiments on various load control algorithms, and year-round testing without the constraints and limitations arising in seasonal field tests with real houses. Experimental results show the practical feasibility of an ensemble of small loads contributing to grid balancing.

Air Conditioners↗

DOE OSTI2022

Frame invariant neural network closures for Kraichnan turbulence

Numerical simulations of geophysical and atmospheric flows have to rely on parameterizations of subgrid scale processes due to their limited spatial resolution. Despite substantial progress in developing parameterization (or closure) models for subgrid scale (SGS) processes using physical insights and mathematical approximations, they remain imperfect and can lead to inaccurate predictions. In recent years, machine learning has been successful in extracting complex patterns from high-resolution spatio-temporal data, leading to improved parameterization models, and ultimately better coarse grid prediction. However, the inability to satisfy known physics and poor generalization hinders the application of these models for real-world problems. In this work, we put forth a frame invariant closure approach to improve the accuracy and generalizability of deep learning-based subgrid scale closure models by embedding physical symmetries directly into the structure of the neural network. Specifically, we utilized specialized layers within the convolutional neural network in such a way that desired constraints are theoretically guaranteed without the need for any regularization terms. We demonstrate our framework for a two-dimensional decaying turbulence test case mostly characterized by the forward enstrophy cascade. We show that our frame invariant SGS model (i) accurately predicts the subgrid scale source term, (ii) respects the physical symmetries such as translation, Galilean, and rotation invariance, and (iii) is numerically stable when implemented in coarse-grid simulation with generalization to different initial conditions and Reynolds number. This work opens up a possibility of connecting physics-based theories and data-driven modeling paradigms, and thus represents a promising step towards the development of physically consistent data-driven turbulence closure models.

42 ENGINEERING↗