SEARCH · Engineering Papers

Results for “Generative models”

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 1,189 records · Page 66

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Firm Synthesizer and Supply-chain Simulator (SynthFirm) v2.0

SynthFirm is a national-scale agent-based freight demand model which generates a complete synthetic population of firms in the U.S. and the business-to-business commodity flows between them. Using publicly available data sources as inputs, SynthFirm simulates detailed firm and fleet characteristics, commodity production and consumption, formation of supply chains, and selection of shipping modes, all of which are essential drivers of commodity flow at a disaggregate level. The SynthFirm 2.0 version includes national commercial vehicle fleet generation, international trade simulation and automized model validation pipeline, which allows seemless deployment across the nation and build a comprehensive freight inventories at national scale or for selected region.

Yang, Hung-Chia [Lawrence Berkeley National Labora↗

DOE OSTI2023

Biophysical Impact of Land-Use and Land-Cover Change on Subgrid Temperature in CMIP6 Models

In this study, we investigate the air temperature response to land-use and land-cover change (LULCC; cropland expansion and deforestation) using subgrid land model output generated by a set of CMIP6 model simulations. Our study is motivated by the fact that ongoing land-use activities are occurring at local scales, typically significantly smaller than the resolvable scale of a grid cell in Earth system models. It aims to explore the potential for a multimodel approach to better characterize LULCC local climatic effects. On an annual scale, the CMIP6 models are in general agreement that croplands are warmer than primary and secondary land (psl; mainly forests, grasslands, and bare ground) in the tropics and cooler in the mid–high latitudes, except for one model. The transition from warming to cooling occurs at approximately 40°N. Although the surface heating potential, which combines albedo and latent heat flux effects, can explain reasonably well the zonal mean latitudinal subgrid temperature variations between crop and psl tiles in the historical simulations, it does not provide a good prediction on subgrid temperature for other land tile configurations (crop vs forest; grass vs forest) under Shared Socioeconomic Pathway 5–8.5 (SSP5–8.5) forcing scenarios. A subset of simulations with the CESM2 model reveals that latitudinal subgrid temperature variation is positively related to variation in net surface shortwave radiation and negatively related to variation in the surface energy redistribution factor, with a dominant role from the latter south of 30°N. We suggest that this emergent relationship can be used to benchmark the performance of land surface parameterizations and for prediction of local temperature response to LULCC.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2020

Global Teleconnections: Atmospheric Interactions Across Scales [Slides]

This project aims at an improved understanding of non-hydrostatic dynamics in atmospheric modeling using the DOE’s next generation climate model. One aspect that we have been focusing on is baroclinic instability which is closely related to mid-latitude cyclone genesis. The other aspect that we recently developed is E3SM simulations under Radiative-Convective Equilibrium (RCE), a prototype of tropical climate.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Development of the IES Plug-and-Play Framework

This report discusses the status of the flexible plug-and-play framework development currently ongoing that aims to integrate Modelica/Dymola with the Risk Analysis and Virtual ENvironment (RAVEN) software in terms of both Functional Mock-Up Interface (FMI)/Functional Mock-Up Unit (FMU) construction and repository structures that aim to ease the sharing and simulation of complex dynamic models. This report aims to provide an overview of all the performed activities resolving around the deployment of methods, software infrastructures, guidelines and workflow for the construction and usage of models, encapsulated using the FMI/FMU protocols and standards. In particular, the report is organized in three main macro-subjects, which are connected to each other: - FMI/FMU adaptors for modelica models - HYBRID repository new structure and open-source deployment - RAVEN FMI/FMU exporting capabilities and Artificial Intelligence (AI)-based analysis acceleration. The first part of the report discusses the FMI/FMU adaptors that have been created within the HYBRID repository to allow users to quickly export models, such as FMUs. Several examples are shown that highlight the step-by-step process of converting an existing Modelica model into an FMU for use within the Dymola platform. Simulation results demonstrate that, while minor differences may occur, the overall control, trends, and solution integrity are maintained between standard Modelica simulation and FMU simulation results. However, it is worth noting that, for small systems, the FMU results have a slower simulation time than the Modelica only simulation. Using this process, a company can provide models that contain proprietary information to entities without disclosing any of the information about the model that could be considered business sensitive. Such an ability would allow institutions to bypass the necessity of “whitewashing” data. In the second part of the report, the new structure of the HYBRID repository is discussed with a major focus on the series of updates that has been completed. These updates include the addition of Modelica system-level regression tests and software quality assurance documentation that ensure that modifications to the Modelica models do not alter system-level model results. The third and final part of the report aims to report the work that has been performed for the deployment of methods and workflows for the construction of RAVEN AI-based models compliant with the FMI/FMU standard. Such development represents the key for the deployment of the concept of “Flexible ecosystem” since it allows for the replacement of high-fidelity modelica models (or any other FMI/FMU compliant model) with RAVEN generated AI surrogate models. Overall, extensive work has been completed on developing FMUs and FMIs from existing models, understanding the requirements and limitations of FMUs, and open-sourcing the HYBRID repository with an integrated regression system.

42 ENGINEERING↗

DOE OSTI2021

Development of the IES Plug-and-Play Framework

14 SOLAR ENERGY↗

DOE OSTI2023

Porting the WAVEWATCH III (v6.07) wave action source terms to GPU

Abstract. Surface gravity waves play a critical role in several processes, including mixing, coastal inundation, and surface fluxes. Despite the growing literature on the importance of ocean surface waves, wind–wave processes have traditionally been excluded from Earth system models (ESMs) due to the high computational costs of running spectral wave models. The development of the Next Generation Ocean Model for the DOE’s (Department of Energy) E3SM (Energy Exascale Earth System Model) Project partly focuses on the inclusion of a wave model, WAVEWATCH III (WW3), into E3SM. WW3, which was originally developed for operational wave forecasting, needs to be computationally less expensive before it can be integrated into ESMs. To accomplish this, we take advantage of heterogeneous architectures at DOE leadership computing facilities and the increasing computing power of general-purpose graphics processing units (GPUs). This paper identifies the wave action source terms, W3SRCEMD, as the most computationally intensive module in WW3 and then accelerates them via GPU. Our experiments on two computing platforms, Kodiak (P100 GPU and Intel(R) Xeon(R) central processing unit, CPU, E5-2695 v4) and Summit (V100 GPU and IBM POWER9 CPU) show respective average speedups of 2× and 4× when mapping one Message Passing Interface (MPI) per GPU. An average speedup of 1.4× was achieved using all 42 CPU cores and 6 GPUs on a Summit node (with 7 MPI ranks per GPU). However, the GPU speedup over the 42 CPU cores remains relatively unchanged (∼ 1.3×) even when using 4 MPI ranks per GPU (24 ranks in total) and 3 MPI ranks per GPU (18 ranks in total). This corresponds to a 35 %–40 % decrease in both simulation time and usage of resources. Due to too many local scalars and arrays in the W3SRCEMD subroutine and the huge WW3 memory requirement, GPU performance is currently limited by the data transfer bandwidth between the CPU and the GPU. Ideally, OpenACC routine directives could be used to further improve performance. However, W3SRCEMD would require significant code refactoring to make this possible. We also discuss how the trade-off between the occupancy, register, and latency affects the GPU performance of WW3.

58 GEOSCIENCES↗

NASA NTRS1984

Reheating the intergalactic medium under extremal conditions

Sherman (1982) commenced the investigation of a hypothetical closure density intergalactic medium (IGM) in the recombination epoch and found present acceptable temperatures in the range from 160,000 to 4,200,000 K for IGM heating. The present investigation has the objective to complete these calculations by finding extreme bounds on the reheating epoch which preceded recombination. The final results are presented in a table. Three models have been generated. The presented models suggest that a closure density IGM is constrained to be heated by a source in the range from 10 to the -31st to 10 to the -27th ergs per s per cu cm. The heating had to originate no earlier than z approximately 28.

Sherman, R. D.↗

NASA NTRS1985

A comparison of precipitating electron energy flux on March 22, 1979 with an empirical model - CDAW 6

Data recorded by Defense Meteorological Satellite Program, Tiros and P-78-1 satellites for the CDAW 6 event on March 22, 1979, have been compared with a statistical model of precipitating electron fluxes. Comparisons have been made on both an orbit-by-orbit basis and on a global basis by sorting and binning the data by AE index, invariant latitude, and magnetic local time in a manner similar to which the model was generated. It is concluded that the model flux agrees with the data to within a factor of two, although small features and the exact locations of features are not consistently reproduced. In addition, the latitude of highest electron precipitation usually occurs about 3 deg more poleward in the model than in the data. This discrepancy is attributed to ring current inflation of the storm time magnetosphere.

Simons, S. L., Jr.↗

NASA NTRS1988

Calculating Robot-Joint Coordinates From Image Coordinates

Detailed knowledge of robot joints not required. Algorithm generates approximate mathematical models of coordinates of joints of robot as functions of coordinates of points in images of work region viewed by television cameras. Joint coordinates necessary to position and orient end effector calculated by mathematical models fitted to experimentally determined data on positions, orientations, and joint coordinates. Generates models as functions of desired location of end effector of robot. Does not require priori knowledge of kinematic equations of robot.

Source record↗

NASA NTRS1994

A toolbox and a record for scientific model development

Scientific computation can benefit from software tools that facilitate construction of computational models, control the application of models, and aid in revising models to handle new situations. Existing environments for scientific programming provide only limited means of handling these tasks. This paper describes a two pronged approach for handling these tasks: (1) designing a 'Model Development Toolbox' that includes a basic set of model constructing operations; and (2) designing a 'Model Development Record' that is automatically generated during model construction. The record is subsequently exploited by tools that control the application of scientific models and revise models to handle new situations. Our two pronged approach is motivated by our belief that the model development toolbox and record should be highly interdependent. In particular, a suitable model development record can be constructed only when models are developed using a well defined set of operations. We expect this research to facilitate rapid development of new scientific computational models, to help ensure appropriate use of such models and to facilitate sharing of such models among working computational scientists. We are testing this approach by extending SIGMA, and existing knowledge-based scientific software design tool.

Ellman, Thomas↗

NASA NTRS1998

Magnetospheric Image Unfolding

The Grant was a three year grant funded under the Space Physics Supporting Research and Technology and Suborbital Program. Our objective was to develop automated techniques needed to unfold or "invert" global images of the magnetospheric ion populations obtained by the new magnetospheric imaging techniques (ENA, EUV) in anticipation of future missions such as the Magnetospheric Imager and, now, IMAGE. Our focus on the present three year grant is to determine the degree to which such images can quantitatively constrain the global electromagnetic properties of the magnetosphere. In a previous three year grant period we successfully automated a forward modeling inversion algorithm, demonstrated that these inversions are robust in the face of realistic instrumental considerations such as counting statistics and backgrounds, applied error analysis techniques to the extracted parameters using variational procedures, implemented very realistic magnetospheric test images to test the inversion algorithms using the Rice University Magnetospheric Specification Model, and began the process of generating parametric models with the flexibility to handle the realistic magnetospheric images (e.g. Roelof et al, 1992; 1993). Our plan for the present 3 year grant period was to complete the development of the inversion tools needed to handle realistic magnetospheric images, assess the degree to which global electrodynamics is quantitatively constrained by ENA images of the magnetosphere, and bring the inversion of EUV images up to the maturity that we will have achieved for the ENA imaging. Below the accomplishments of our three year effort are present followed by a list of our presentations and publications. The accomplishments of all three years are presented here, and thus some of these items appeared on interim progress reports.

Source record↗

NASA NTRS2011

Automated Verification of Specifications with Typestates and Access Permissions

We propose an approach to formally verify Plural specifications based on access permissions and typestates, by model-checking automatically generated abstract state-machines. Our exhaustive approach captures all the possible behaviors of abstract concurrent programs implementing the specification. We describe the formal methodology employed by our technique and provide an example as proof of concept for the state-machine construction rules. The implementation of a fully automated algorithm to generate and verify models, currently underway, provides model checking support for the Plural tool, which currently supports only program verification via data flow analysis (DFA).

Siminiceanu, Radu I.↗

NASA NTRS2012

Modeling of Damage Initiation and Progression in a SiC/SiC Woven Ceramic Matrix Composite

The goal of an ongoing project at NASA Glenn is to investigate the effects of the complex microstructure of a woven ceramic matrix composite and its variability on the effective properties and the durability of the material. Detailed analysis of these complex microstructures may provide clues for the material scientists who `design the material? or to structural analysts and designers who `design with the material? regarding damage initiation and damage propagation. A model material system, specifically a five-harness satin weave architecture CVI SiC/SiC composite composed of Sylramic-iBN fibers and a SiC matrix, has been analyzed. Specimens of the material were serially sectioned and polished to capture the detailed images of fiber tows, matrix and porosity. Open source analysis tools were used to isolate various constituents and finite elements models were then generated from simplified models of those images. Detailed finite element analyses were performed that examine how the variability in the local microstructure affected the macroscopic behavior as well as the local damage initiation and progression. Results indicate that the locations where damage initiated and propagated is linked to specific microstructural features.

Mital, Subodh K.↗

NASA NTRSDate not supplied

Simulation of Dynamic Crushing Behavior of Simulated Carbon/Epoxy Stanchions Using MAT58 and MAT213 in LS-DYNA

Carbon fiber-reinforced polymer (CFRP) composites have shown promise as a material for structures designed to absorb energy in crush-style impact loading. In complementing the lightweight and tailorable characteristics of CFRP structures, the benefits of impact energy absorption are particularly interesting in aerospace vehicles that require excellent crashworthiness performance. However, simulating the behavior of CFRP structures in dynamic, crushing-style impact loading is challenging because of the many damage and failure modes that are essential to capture in the model. For the present investigation, CFRP stanchions were tested using a crash sled experimental test rig. The stanchions were designed by the Composite Materials Handbook-17 (CMH-17) Crashworthiness Working Group for the purpose of comparing experimental crash sled tests to corresponding simulations of the tests. The stanchions are C-channel shaped and represent a geometry common in the interior of aerospace vehicle structures. Explicit simulations in LSDYNA were performed using a well-established composite material model (MAT58) and a next-generation material model (MAT213). Fully integrated shell elements were utilized instead of 3D solid elements to limit computation time. In all models, each of the sixteen plies were represented with individual layers of shell elements with tiebreak contact between each adjacent ply pair to simulate interlaminar fracture. Simulating the crushing of the stanchions occurred in two phases. First, the material models were calibrated using flat specimens that were manufactured with the same layup as the stanchions. While the original goal was calibration of material-related properties, meshdependent behavior was observed in simulations with either material model, and an unstructured mesh was selected to remediate undesirable mesh-dependent failure modes. Additionally, for both the MAT58 and MAT213 models, it was found that either the crush force or the failure mode could be modeled accurately, but no set of parameters could be identified to attain both results in the same model. Once satisfactory calibration was achieved, the same material parameters were applied to the stanchion crush simulations. The stanchion simulations showed that MAT213 more accurately predicted the experimentally determined crush force, and both material models predicted key aspects of the experimentally observed failure modes.

Composite Materials↗

DOE OSTI2022

Language models for the prediction of SARS-CoV-2 inhibitors

The COVID-19 pandemic highlights the need for computational tools to automate and accelerate drug design for novel protein targets. We leverage deep learning language models to generate and score drug candidates based on predicted protein binding affinity. We pre-trained a deep learning language model (BERT) on ∼9.6 billion molecules and achieved peak performance of 603 petaflops in mixed precision. Our work reduces pre-training time from days to hours, compared to previous efforts with this architecture, while also increasing the dataset size by nearly an order of magnitude. For scoring, we fine-tuned the language model using an assembled set of thousands of protein targets with binding affinity data and searched for inhibitors of specific protein targets, SARS-CoV-2 Mpro and PLpro. We utilized a genetic algorithm approach for finding optimal candidates using the generation and scoring capabilities of the language model. Our generalizable models accelerate the identification of inhibitors for emerging therapeutic targets.

Blanchard, Andrew E.↗

NASA NTRS1990

Analysis and design of a high power, digitally-controlled spacecraft power system

The progress to date on the analysis and design of a high power, digitally controlled spacecraft power system is described. Several battery discharger topologies were compared for use in the space platform application. Updated information has been provided on the battery voltage specification. Initially it was thought to be in the 30 to 40 V range. It is now specified to be 53 V to 84 V. This eliminated the tapped-boost and the current-fed auto-transformer converters from consideration. After consultations with NASA, it was decided to trade-off the following topologies: (1) boost converter; (2) multi-module, multi-phase boost converter; and (3) voltage-fed push-pull with auto-transformer. A non-linear design optimization software tool was employed to facilitate an objective comparison. Non-linear design optimization insures that the best design of each topology is compared. The results indicate that a four-module, boost converter with each module operating 90 degrees out of phase is the optimum converter for the space platform. Large-signal and small-signal models were generated for the shunt, charger, discharger, battery, and the mode controller. The models were first tested individually according to the space platform power system specifications supplied by NASA. The effect of battery voltage imbalance on parallel dischargers was investigated with respect to dc and small-signal responses. Similarly, the effects of paralleling dischargers and chargers were also investigated. A solar array and shunt model was included in these simulations. A model for the bus mode controller (power control unit) was also developed to interface the Orbital replacement Unit (ORU) model to the platform power system. Small signal models were used to generate the bus impedance plots in the various operating modes. The large signal models were integrated into a system model, and time domain simulations were performed to verify bus regulation during mode transitions. Some changes have subsequently been incorporated into the models. The changes include the use of a four module boost discharger, and a new model for the mode controller, which includes the effects of saturation. The new simulations for the boost discharger show the improvement in bus ripple that can be achieved by phase-shifted operation of each of the boost modules.

Lee, F. C.↗

DOE OSTI2022

Synthetic data generation for machine learning model training for energy theft scenarios using cosimulation

Abstract Technical and non‐technical losses in distribution circuits result in significant economic costs to power utilities. One type of non‐technical loss is energy theft by various means including illegal tapping of feeders, bypassing the meter, and billing fraud. These losses are usually hard to detect, and can remain undetected for long periods of time. Machine learning models have been proven effective in detecting these conditions, but rely on the availability of large, good‐quality training data sets. The problem is exacerbated by the imbalanced nature of data related to these conditions—energy theft, though costly, is very rare. The available data sets generally have very few samples of theft with most of the data pertaining to normal operation. Such data sets are generally not suitable to train machine learning models. In this paper, an overview of energy theft detection techniques, the challenges with their data needs, and the limitations of current techniques to bridge such data limitations is presented. A co‐simulation framework is proposed to generate reliable training data for machine learning algorithms for theft detection. An example scenario is presented and a machine learning model is built to detect certain kinds of energy theft.

24 POWER TRANSMISSION AND DISTRIBUTION↗

NASA NTRS1976

A method for generating numerical pilot opinion ratings using the optimal pilot model

A method for generating numerical pilot opinion ratings using the optimal pilot model is introduced. The method is contained in a rating hypothesis which states that the numerical rating which a human pilot assigns to a specific vehicle and task can be directly related to the numerical value of the index of performance resulting from the optimal pilot modeling procedure as applied to that vehicle and task. The hypothesis is tested using the data from four piloted simulations. The results indicate that the hypothesis is reasonable, but that the predictive capability of the method is a strong function of the accuracy of the pilot model itself. This accuracy is, in turn, dependent upon the parameters which define the optimal modeling problem. A procedure for specifying the parameters for the optimal pilot model in the absence of experimental data is suggested.

Hess, R. A.↗