Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “COLLECTIVE 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 793 records · Page 44
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

Aero-Propulsive Modeling for eVTOL Aircraft Using Wind Tunnel Testing with Multisine Inputs

A novel approach for modeling the aero-propulsive characteristics of an electric vertical takeoff and landing (eVTOL) aircraft was developed and demonstrated in wind tunnel testing. The approach was applied to the NASA LA-8 tandem tilt-wing eVTOL aircraft, using an efficient, hybrid experiment design composed of a static I-optimal response surface design for slowly-varying test variables, and dynamic orthogonal phase-optimized multisine excitations for the control surfaces and electric propulsors. Both the static and dynamic experiment designs were executed simultaneously to collect informative data for model identification. Statistically-weighted multivariate orthogonal function modeling was used to combine local modeling results computed in the frequency domain using data collected with dynamic excitations operating on the control effectors to form an aggregate aero-propulsive model. The final identified model exhibited good predictive capability when compared to validation data acquired separately from the data used to develop the model. The required test time using these new techniques was reduced by at least a factor of five compared to previous static wind tunnel testing for the LA-8 aircraft, while providing more informative data, greater parameterization flexibility, and high-quality models.

system identification↗
DOE OSTI2012

Financial and economic determinants of collective action: The case of wastewater management

Where public environmental funds support development of wastewater infrastructure, funding institutions ensure the economic use of funds, while the beneficiaries minimize their own costs. In rural areas, there is often a choice between decentralized or centralized (multi-village) systems: if the centralized system is most economic, then only this system is eligible for public funding. However, its implementation requires a voluntary cooperation of the concerned communities, who need to organize themselves to develop and run the infrastructure. The paper analyzes the social determinants of collaboration in a generic case study, using the following variables: method of (economic) assessment, modeled by the social discount rate, funding policy, modeled by the funding rate, and users' self-organization, modeled by cost sharing. In a borderline situation, where the centralized system turns out to be most economic, but this assessment is contingent on the assessment method, collective action may fail: the advantages of collective action from funding are too small to outweigh organizational deficiencies. Considering in this situation sanitation as a human right, authors recommend using innovative forms of organization and, if these fail, reassessing either the amount of funding or the eligibility for funding of more acceptable alternatives. - Highlights: Black-Right-Pointing-Pointer A generic case study models collective action and funding in wastewater management. Black-Right-Pointing-Pointer Determinants of success: economic assessment, funding policy and self-organization. Black-Right-Pointing-Pointer Success indicators: conflict rate, funds needed to make cost shares fair. Black-Right-Pointing-Pointer Method for analyzing centralized vs. decentralized disputes. Black-Right-Pointing-Pointer If collective action has less benefits, innovative cost sharing may ensure success.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Evaluation of Deep Subsurface Resistivity Imaging for Hydrofracture Monitoring

This report describes the results of the first of its kind monitoring of a hydrofracture operation with electromagnetic measurements. The researchers teamed with oil and gas producer Encana Corporation to design and execute a borehole to surface monitoring of three fracture stages at a well pad in central Colorado. The field project consisted of an equipment upgrade, a survey design and modeling phase, several weeks of data collection, and data processing and interpretation. Existing Depth to Surface Resistivity (DSR) instrumentation was upgraded to allow for continuous high precision recording from downhole sources. The full system can now collect data continuously for up to 72 hours, which is sufficient to measure data for 10 frac stages. Next we used numerical modeling and frac treatment data supplied by Encana to design a field survey to detect EM signal from induced fractures. Prior to modeling we developed a novel technique for using well casing as an antenna for a downhole source. Modeling shows that 1) a measurable response for an induced fracture could be achieved if the facture fluid was of high salinity 2) an optimum fracture response is created when the primary source field is parallel to the well casing but perpendicular to the fracture direction. In mid-July, 2014 we installed an array of more than 100 surface sensors, distributed above the treatment wells and extending for approximately 1 km north and 750 m eastward. We applied a 0.6 Hz square wave signal to a downhole current electrode located in a horizontal well 200 m offset from the treatment well with a return electrode on the surface. The data were transmitted to a recording trailer via Wi-Fi where we monitored receiver and transmitter channels continuously in a 72-hour period which covered 7 frac stages, three of which were high salinity. Although the background conditions were very noisy we were able to extract a clear signal from the high salinity stages. Initial data interpretation attempts consisting of trying to directly invert collected data using a simple background starting model did not produce reasonable models. We next used a simulated data set to develop a constrained inversion workflow that places the downhole fracture in the correct location. Finally, we used a combination of forward and inverse models to fit the collected data to a model that incorporated 3 frac stages. This project provided a first-of-its-kind application of EM technology to map an induced fracture. The results demonstrated a clear anomaly during the operation that can be fit to a reasonable model of an induced fracture.

42 ENGINEERING↗
DOE OSTI2016

Aggregate modeling of fast-acting demand response and control under real-time pricing

This paper develops and assesses the performance of a short-term demand response (DR) model for utility load control with applications to resource planning and control design. Long term response models tend to underestimate short-term demand response when induced by prices. This has two important consequences. First, planning studies tend to undervalue DR and often overlook its benefits in utility demand management program development. Second, when DR is not overlooked, the open-loop DR control gain estimate may be too low. This can result in overuse of load resources, control instability and excessive price volatility. Our objective is therefore to develop a more accurate and better performing short-term demand response model. We construct the model from first principles about the nature of thermostatic load control and show that the resulting formulation corresponds exactly to the Random Utility Model employed in economics to study consumer choice. The model is tested against empirical data collected from field demonstration projects and is shown to perform better than alternative models commonly used to forecast demand in normal operating conditions. The results suggest that (1) existing utility tariffs appear to be inadequate to incentivize demand response, particularly in the presence of high renewables, and (2) existing load control systems run the risk of becoming unstable if utilities close the loop on real-time prices.

Aggregate Modeling↗
DOE OSTI2022

Contextual Active Online Model Selection with Expert Advice

How can we collect the most useful labels to learn a model selection policy, when presented with arbitrary heterogeneous data streams? In this paper, we formulate this task as a contextual active model selection problem, where at each round the learner receives an unlabeled data point along with a context. The goal is to output the best model for any given context without obtaining an excessive amount of labels. In particular, we focus on the task of selecting pre-trained classifiers, and propose a contextual active model selection algorithm (CAMS), which relies on a novel uncertainty sampling query criterion defined on a given policy class for adaptive model selection. In comparison to prior art, our algorithm does not assume a globally optimal model. We provide rigorous theoretical analysis for the regret and query complexity under both adversarial and stochastic settings. Our experiments on several benchmark classification datasets demonstrate the algorithm’s effectiveness in terms of both regret and query complexity. Notably, to achieve the same accuracy, CAMS incurs less than 10% of the label cost when compared to the best online model selection baselines on CIFAR10.

Liu, Xuefeng↗
DOE OSTI2024

Formulation of a one-dimensional electrostatic plasma model for testing the validity of kinetic theory

Here, we present a one-dimensional (1-D) model composed of aligned, electrostatically interacting charged disks, conceived to address in a computable model the validity of the Bogoliubov assumption on the decay of particle correlations in the Born–Bogoliubov–Green–Kirkwood–Yvon hierarchy. This assumption is a basic premise of plasma kinetic theory. The disk model exhibits spatially 1-D features at short distances, but retains 3-D features at large distances. Here the collective dynamics of this model plasma is investigated by solving the corresponding Vlasov equation. In addition, the implementation of the model for the numerical validation of the Bogoliubov assumption is formulated.

1-D plasma model↗
DOE OSTI2026

Surface ozone and meteorological variables at CoURAGE TBS site during summer IOP

This dataset provides continuous near ground-level in situ ozone and meteorological data at the TBS site (S7) during the CoURAGE summer IOP. The ozone inlet and the weather station were approximately 5 meters above the surface. In situ ozone (ppbv) was collected with a 2BTech Model 205 dual beam ozone monitor. The meteorological data (temperature, pressure, relative humidity, wind speed, and wind direction) was collected with an AIRMAR 220WX WeatherStation Instrument with relative humidity (RH) module. The data has 2-second temporal resolution.

bar_pres↗
DOE OSTI2018

Rapidly moving new bacteria to model-organism status

The paradigm of large research communities collectively working on a small number of model bacteria such as Escherichia coli and Bacillus subtilis is changing. While these classic model bacteria will continue to be important for advanced systems biology and new technology development, we envision that increasingly small research teams will be deeply investigating their own favorite strains, for example as new hosts for metabolic engineering or as key members of a complex microbiome. Given the lack of a research community and the sheer number of possible bacteria to interrogate, the development and application of technologies to rapidly and inexpensively advance these unstudied strains to ‘model-organism’ status is imperative. Here, we discuss the minimal information and tools necessary to develop a new model bacterium and how existing approaches can bring this power into the hands of a single investigator.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Stochastic Virtual Battery Modeling of Uncertain Electrical Loads using Variational Autoencoder

Effective utilization of flexible loads for grid services, while satisfying end-user preferences and constraints, requires an accurate estimation of the aggregated predictive flexibility offered by the electrical loads. Recently, there have been efforts to quantify the predictive flexibility of thermostatic loads (e.g. residential air-conditioners, electric water-heaters) using the notion of virtual battery (VB), whose state evolution is governed by a first order dynamics including self-dissipation rate, and power and energy capacities. Identifying the VB model parameters for a collection of thermostatic loads, however, is challenging primarily due to uncertainties and lack of information regarding the end-user behavior, underlying device models and parameters. In this paper, we propose a \textit{variational autoencoder}-based deep learning algorithm to identify the parameters of the VB model. Using available sensors and meters data, the proposed algorithm generates not only point estimates of the VB parameters, but also confidence intervals around those values. Effectiveness of the proposed frameworks is demonstrated on a collection of electric water-heater loads, whose operation is driven by uncertain water usage profiles.

virtual battery, deep learning algorithms↗
DOE OSTI2017

Soil-mediated effects of global change on plant communities depend on plant growth form

Understanding why species respond to climate change is critical for forecasting invasions, diversity, and productivity of communities. Although researchers often predict species’ distributions and productivity based on direct physiological responses to environments, theory suggests that striking shifts in community composition could arise if global change alters indirect feedbacks mediated by resources, mutualists, or antagonists. To test whether global change influences plant communities via soil–mediated feedbacks, we grew model communities in soils collected from a seven–year field manipulation of CO 2 , warming, and invasion. We evaluated mechanisms underlying variation in the model communities by comparing species’ growth in equivalent soil histories with, and without, experimentally reduced soil biota (via sterilization) and nutrient limitation (via fertilization). We show that grasses performed consistently across all soil history scenarios and that soil biota limited grasses more than nutrients. In contrast, forbs were differentially sensitive to soil history scenarios, with the magnitude and direction of responses to soil biota and nutrients dependent upon plant species and global change scenario. The asymmetry in importance of soil history for grasses and forbs is likely explained by differences in life history strategy. We conclude that accounting for species’ growth strategies will improve predictions of species sensitivity to altered soil feedbacks in future climates.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2019

Thermal conductivity measurements and modeling of ceramic fiber insulation materials

Ceramic fiber insulation materials are used in numerous applications (e.g. aerospace, fire protection, and military) for their stability and performance in extreme environments. However, the thermal properties of these materials have not been thoroughly characterized for many of the conditions that they will be exposed to, such as high temperatures, pressures, and alternate gaseous atmospheres. The resulting uncertainty in the material properties can complicate the design of systems using these materials. In this study, the thermal conductivity of two ceramic fiber insulations, Fiberfrax T-30LR laminate and 970-H paper, was measured as a function of atmospheric temperature and compression in an air environment using the transient plane source technique. Furthermore, a model is introduced to account for changes in thermal conductivity with temperature, compression, and ambient gas. The model was tuned to the collected experimental data and results are compared. Lastly, the tuned model is also compared to published data sets taken in argon, helium, and hydrogen environments and agreement is discussed.

36 MATERIALS SCIENCE↗
DOE OSTI2010

Structure of the doublet bands in doubly odd nuclei: The case of {sup 128}Cs

The structure of the {Delta}J=1 doublet bands in {sup 128}Cs is investigated within the framework of the interacting vector boson-fermion model. A new, purely collective interpretation of these bands is given on the basis of the used boson-fermion dynamical symmetry of the model. The energy levels of the doublet bands as well as the absolute B(E2) and B(M1) transition probabilities between the states of both yrast and yrare bands are described quite well. The observed odd-even staggering of both B(M1) and B(E2) values is reproduced by the introduction of an appropriate interaction term of quadrupole type, which produces such a staggering effect in the transition strengths. The calculations show that the appearance of doublet bands in certain odd-odd nuclei could be a consequence of the realization of a larger dynamical symmetry based on the noncompact supersymmetry group OSp(2{Omega}/12,R).

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2013

C6 plate puncture testing report.

There are numerous scenarios where critical systems could be subject to penetration by projectiles or fixed objects (e.g., collision, natural disaster, act of terrorism, etc.). It is desired to use computational models to examine these scenarios and make risk-informed decisions; however, modeling of material failure is an active area of research, and new models must be validated with experimental data. The purpose of this report is to document the experimental work performed from FY07 through FY08 on the Campaign Six Plate Puncture project. The goal of this project was to acquire experimental data on the puncture and penetration of metal plates for use in model validation. Of particular interest is the PLH failure model also known as the multilinear line segment model. A significant amount of data that will be useful for the verification and validation of computational models of ductile failure were collected during this project were collected and documented herein; however, much more work remains to be performed, collecting additional experimental data that will further the task of model verification.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2026

Enhancement of HLW glass property-composition models

Since the WTP compositional space of HLW glasses is extremely large, development of HLW property-composition models is a multi-year task consisting of multiple phases. The primary focus of this work was to enhance the WTP HLW models of interest including PCT releases, spinel crystallization (T1%), viscosity, electrical conductivity, and TCLP-Cd response. Model development to predict nepheline formation upon CCC is the subject of a separate task. In particular, the earlier work has produced property-composition models for glass melt viscosity and glass melt electrical conductivity that showed good performance, while models for PCT releases and spinel crystallization (T1%) required improvement. Therefore, more efforts were directed in the present work to collect data and improve model performance for HLW glass PCT releases and spinel crystallization than for other properties. The present work is a continuation of earlier development phases and is responsive to the applicable Test Plan.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2019

DeepWelding: a Deep Learning Enhanced Approach to GTAW Using Multi-source Sensing Images

Deep learning has great potential to reshape manufacturing industries. In this paper, we present DeepWelding, a novel framework that applies deep learning techniques to improve gas tungsten arc welding (GTAW) process monitoring and penetration detection using multi-source sensing images. The framework is capable of analyzing multiple types of optical sensing images synchronously and consists of three deep learning enhanced consecutive phases: image preprocessing, image selection, and weld penetration classification. Specifically, we adopted generative adversarial networks (pix2pix) for image denoising and classic convolutional neural networks (AlexNet) for image selection. Both pix2pix and AlexNet delivered satisfactory performance. However, five individual neural networks with heterogeneous architectures demonstrated inconsistent generalization capabilities in the classification phase when holding out multi-source images generated with specific experimental settings. Therefore, two ensemble methods combining multiple neural networks were designed to improve the model performance on unseen data collected from different experimental settings. In conclusion, we also found that the quality of model prediction was heavily influenced by the data stream collection environment. We think these findings are beneficial for the broad intelligent welding community.

42 ENGINEERING↗
DOE OSTI2021

Carbon–Nutrient Economy of the Rhizosphere: Improving Biogeochemical Prediction and Scaling Feedbacks from Ecosystem to Regional Scales

Our project has advanced the science of plant-soil-microbial dynamics across these areas: i) nutrient cycling and plant uptake; ii) root exudation and priming; and, iii) mycorrhizal dynamics. Our project has accomplished 5 main developments: 1) Incorporation of phosphorus cycling into the Fixation & Uptake of Nutrients (FUN 3.0) model. 2) Coupling of FUN 3.0 into the E3SM Land Model (ELM). 3) Data collection across a large mycorrhizal gradient in the US, as well data in the tropics, to parameterize, test, and validate the model. 4)Scaling up mycorrhizal association measurements across landscapes using airborne hyperspectral remote sensing data. 5) Evaluation of global carbon and nutrient cycle impacts in the Community Land Model (CLM5.0) from a suite of new global mycorrhizal association maps. Over 25 publications resulted from this project, with more continuing past the project funded lifetime. Paper highlights from most of these publications have already been submitted to the DOE paper submission online system. These publications include journals such as Science and PNAS, as well as top disciplinary journals from the Nature journals, Global Change Biology, New Phytologist, and Ecology Letters, for example. Our project also contributed to improving the process representation, capabilities, and accuracy of the DOE ELM. Overall, this project significantly advanced the science of belowground plant-soil-microbial interactions as well as technical capabilities from remote sensing to modeling.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

torch-einshard v1.0

torch-einshard is a Python library for describing local and distributed PyTorch tensor computations with compact, einsum-like notation. Its expressions name logical axes, specify how they are sharded across a PyTorch DeviceMesh, and represent partial reductions. The library automatically performs contractions, permutations, reshaping, splitting, gathering, reduction, reduce-scatter, and repartitioning while preserving autograd. Additional features include sharding-aware FFTs, tensor rolls, halo exchange, sliding windows, 1D–3D convolutions, uneven-shard handling, parameter initialization and gradient management, and cost-based execution planning. It is designed for scientific machine learning and large-model workloads, including tensor-, sequence-, and spatial-parallel MLPs, attention, convolutions, and spectral operations. Compared with manually combining torch.einsum and distributed collectives, torch-einshard expresses both the mathematical operation and data placement in one readable formula. This reduces boilerplate and synchronization errors, keeps forward and backward communication consistent, and allows the library to select optimized collective strategies without changing model code.

Morozov, Dmitriy [Lawrence Berkeley National Labor↗