Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “representation”

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 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Optimal Network Topology for Node-Breaker Representations With AC Power Flow Constraints

It has been demonstrated that network topology optimization (NTO) may change the topology of power system networks, and consequently, provide additional flexibility to reduce network congestion and violations. Most NTO problems are formulated based on the bus-branch model in which it is challenging to represent a realistic picture of all substation configurations. In this paper, we explore advantages of substation reconfiguration modeling based on node-breaker representations for NTO problem with full nonlinear alternating current power flow. It also proposes a tailored solution algorithm to solve this nonconvex mixed-integer nonlinear programming through the outer approximation method. The proposed solution approach iterates between a mixed-integer linear programming and a nonlinear subproblem. Additional enhancements to further accelerate the iteration process are illustrated. Numerical case studies demonstrate the relative economic and operational impact of optimal network topology with node-breaker representations.

42 ENGINEERING↗
DOE OSTI2024

Efficient graph representation framework for chemical molecule similarity tasks

Graph data has emerged in numerous scientific domains and machine learning techniques have been widely used for analysis and learning of diverse data for prediction and decision. Machine learning techniques can readily address complex problems by leveraging their structural information. But graphs cannot be directly used for existing machine learning algorithms unless encoded as vectors. The problem of efficient representation of graphs is a substantial challenge in graph machine learning. In this paper, we propose a novel two-stage framework for the representation of chemical molecule graphs based on the strengths of Graph Isomorphism Networks (GINs) and Siamese autoencoders. In the first stage, the GIN model is constructed and trained using the structural information of chemical molecule graphs. Node attributes, edge attributes, and edge indices are used as input data, while graph attributes are used as labels. The GIN model effectively captures the structural characteristics of graphs and can accurately predict graph attributes, i.e., molecular properties. It also generates Graph Embeddings, represented as vectors that encode the structural information of graphs. In the second stage, Graph Embedding vectors are further optimized for downstream similarity tasks while preserving the graph structural information. The Siamese autoencoder is constructed and trained, which reduces the dimensionality of the Graph Embedding vectors, while maximizing the preservation of structural information in the original high-dimensional vectors. The resulting low-dimensional Graph Embeddings can be effectively utilized for tasks such as approximate nearest neighbor search. The experimental results demonstrate the effectiveness of our proposed framework in accurately predicting graph similarity.

Ma, Jiaji↗
DOE OSTI2022

Retargetable Optimizing Compilers for Quantum Accelerators via a Multi-Level Intermediate Representation

In this work, we present a multi-level quantum-classical intermediate representation (IR) that enables an optimizing, retargetable compiler for available quantum languages. Our work builds upon the Multi-level Intermediate Representation (MLIR) framework and leverages its unique progressive lowering capabilities to map quantum languages to the LLVM machine-level IR. We provide both quantum and classical optimizations via the MLIR pattern rewriting sub-system and standard LLVM optimization passes, and demonstrate the programmability, compilation, and execution of our approach via standard benchmarks and test cases. In comparison to other standalone language and compiler efforts available today, our work results in compile times that are 1000x faster than standard Pythonic approaches, and 5-10x faster than comparative standalone quantum language compilers. Our compiler provides quantum resource optimizations via standard programming patterns that result in a 10x reduction in entangling operations, a common source of program noise. We see this work as a vehicle for rapid quantum compiler prototyping.

43 PARTICLE ACCELERATORS↗
DOE OSTI2026

REV-INR: Regularized Evidential Implicit Neural Representation for Uncertainty-Aware Volume Visualization

Applications of Implicit Neural Representations (INRs) have emerged as a promising deep learning approach for compactly representing large volumetric datasets. These models can act as surrogates for volume data, enabling efficient storage and on-demand reconstruction via model predictions. However, conventional deterministic INRs only provide value predictions without insights into the model’s prediction uncertainty or the impact of inherent noisiness in the data. This limitation can lead to unreliable data interpretation and visualization due to prediction inaccuracies in the reconstructed volume. Identifying erroneous results extracted from model-predicted data may be infeasible, as raw data may be unavailable due to its large size. To address this challenge, we introduce REV-INR, Regularized Evidential Implicit Neural Representation, which learns to predict data values accurately along with the associated coordinate-level data uncertainty and model uncertainty using only a single forward pass of the trained REV-INR during inference. By comprehensively comparing and contrasting REV-INR with existing well-established deep uncertainty estimation methods, we show that REV-INR achieves the best volume reconstruction quality with robust data (aleatoric) and model (epistemic) uncertainty estimates using the fastest inference time. Consequently, we demonstrate that REV-INR facilitates assessment of the reliability and trustworthiness of the extracted isosurfaces and volume visualization results, enabling analyses to be solely driven by model-predicted data.

Saklani, Shanu [Indian Institute of Technology, Ka↗
DOE OSTI2025

Distributed Stochastic Optimization of a Neural Representation Network for Time-Space Tomography Reconstruction

4D time-space reconstruction of dynamic events or deforming objects using X-ray computed tomography (CT) is an important inverse problem in non-destructive evaluation. Conventional back-projection based reconstruction methods assume that the object remains static for the duration of several tens or hundreds of X-ray projection measurement images (reconstruction of consecutive limited-angle CT scans). However, this is an unrealistic assumption for many in-situ experiments that causes spurious artifacts and inaccurate morphological reconstructions of the object. To solve this problem, we propose to perform a 4D time-space reconstruction using a distributed implicit neural representation (DINR) network that is trained using a novel distributed stochastic training algorithm. Our DINR network learns to reconstruct the object at its output by iterative optimization of its network parameters such that the measured projection images best match the output of the CT forward measurement model. Here, we use a forward measurement model that is a function of the DINR outputs at a sparsely sampled set of continuous valued 4D object coordinates. Unlike previous neural representation architectures that forward and back propagate through dense voxel grids that sample the object's entire time-space coordinates, we only propagate through the DINR at a small subset of object coordinates in each iteration resulting in an order-of-magnitude reduction in memory and compute for training. DINR leverages distributed computation across several compute nodes and GPUs to produce high-fidelity 4D time-space reconstructions. We use both simulated parallel-beam and experimental cone-beam X-ray CT datasets to demonstrate the superior performance of our approach.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Distributed Neural Representation for Reactive In Situ Visualization

Implicit neural representations (INRs) have emerged as a powerful tool for compressing large-scale volume data. This opens up new possibilities for in situ visualization. However, the efficient application of INRs to distributed data remains an underexplored area. Here, in this work, we develop a distributed volumetric neural representation and optimize it for in situ visualization. Our technique eliminates data exchanges between processes, achieving state-of-the-art compression speed, quality and ratios. Our technique also enables the implementation of an efficient strategy for caching large-scale simulation data in high temporal frequencies, further facilitating the use of reactive in situ visualization in a wider range of scientific problems. We integrate this system with the Ascent infrastructure and evaluate its performance and usability using real-world simulations.

Wu, Qi↗
DOE OSTI2020

A Numerical Study of Hypoxia in Chesapeake Bay Using an Unstructured Grid Model: Validation and Sensitivity to Bathymetry Representation

A three-dimensional unstructured-grid hydrodynamic and water quality model (Semi-implicit Cross-scale Hydroscience Intergrated System Model-Integrated Compartment Model) is applied successfully for Chesapeake Bay. The model is validated with observations of salinity, chlorophyll-a, dissolved oxygen, nutrients, and phytoplankton productions from the year 1991 to 1995 for the mainstem and some major tributaries, based on multiple model skill scores. Model experiments are conducted to test the importance of having (1) an accurate representation of bathymetry to correctly predict hypoxia and other processes and (2) a high-resolution model grid for tributaries to correctly simulate water quality variables. Further, comparison with the model experiment results with bathymetry smoothing indicates that bathymetry smoothing, as commonly used for many systems, changes the stratification and lateral circulation pattern, resulting in more salt intrusion into shallow water regions, and an increase in the freshwater age. Consequently, a model with bathymetry smoothing can lead to an unrealistic prediction of the distribution of hypoxia and phytoplankton production. Local grid refinement shows significant improvement of model simulations on local stratification and water quality variables. Overall, the use of high-resolution unstructured grid model leads to a faithful representation of the complex geometry, and thus a seamless cross-scale capability for simulating water quality processes in the Bay including tributaries and tidal creeks.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Lineage-based functional types: characterising functional diversity to enhance the representation of ecological behaviour in Land Surface Models

Process-based vegetation models attempt to represent the wide range of trait variation in biomes by grouping ecologically similar species into plant functional types (PFTs). This approach has been successful in representing many aspects of plant physiology and biophysics but struggles to capture biogeographic history and ecological dynamics that determine biome boundaries and plant distributions. Grass-dominated ecosystems are broadly distributed across all vegetated continents and harbour large functional diversity, yet most Land Surface Models (LSMs) summarise grasses into two generic PFTs based primarily on differences between temperate C 3 grasses and (sub)tropical C 4 grasses. Incorporation of species-level trait variation is an active area of research to enhance the ecological realism of PFTs, which form the basis for vegetation processes and dynamics in LSMs. Using reported measurements, we developed grass functional trait values (physiological, structural, biochemical, anatomical, phenological, and disturbance-related) of dominant lineages to improve LSM representations. In this work, our method is fundamentally different from previous efforts, as it uses phylogenetic relatedness to create lineage-based functional types (LFTs), situated between species-level trait data and PFT-level abstractions, thus providing a realistic representation of functional diversity and opening the door to the development of new vegetation models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Software For Detecting, Analyzing, And Mitigating Anomalous Network Packets Through A Multi Dimensional Latent Space Representation

The code ingests network packet payloads, classifies the payload as anomalous or normal, and plots the payload in a 3-D latent space representation. It also enables user interaction with the representation to allow for the user to click on the packet payload in the latent space and determine host id, network topology, and access the raw packet data.

Anderson, MatthewW.↗
DOE OSTI2023

HD-Bind: Encoding of Molecular Structure with Low Precision, Hyperdimensional Binary Representations

Publicly available collections of drug-like molecules have grown to comprise tens of billions of compounds due to advances in combinatorial chemistry. Traditional methods for identifying "hit" molecules from a large collection of potential drug-like candidates have relied on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have the major drawback that they require exceptional computing capabilities for even relatively small collections of molecules. Hyperdimensional Computing (HDC) is a recently-proposed learning paradigm that represents data with high-dimension binary vectors; this allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas. We consider existing HDC approaches for molecular property classification and introduce two novel encodings of a commonly-used molecular representation, the extended connectivity fingerprint (ECFP). We show that HDC-based inference methods are as much as 91 times more efficient than traditional machine learning methods, and achieve an acceleration of nearly nine orders of magnitude compared to molecular docking. Our results show that HDC accelerated methods retain competitive accuracy on a number of well-studied tasks such as molecular property predictions using the MoleculeNet dataset, and bind/no-bind activity classification using the DUD-E and LIT-PCBA datasets. Our work thus motivates further investigation into molecular representation learning to develop ultraefficient pre-screening tools.

Jones, William↗
DOE OSTI2020

Representation of Modes of Variability in Six U.S. Climate Models

We compare the performance of several modes of variability across six U.S. climate modeling groups, with a focus on identifying robust improvements in recent models [including those participating in phase 6 of the Coupled Model Intercomparison Project (CMIP)] compared to previous versions. In particular, we examine the representation of the Madden–Julian oscillation (MJO), El Niño–Southern Oscillation (ENSO), the Pacific decadal oscillation (PDO), the quasi-biennial oscillation (QBO) in the tropical stratosphere, and the dominant modes of extratropical variability, including the southern annular mode (SAM), the northern annular mode (NAM) [and the closely related North Atlantic Oscillation (NAO)], and the Pacific–North American pattern (PNA). Where feasible, we explore the processes driving these improvements through the use of “intermediary” experiments that utilize model versions between CMIP3/5 and CMIP6 as well as targeted sensitivity experiments in which individual modeling parameters are altered. We find clear and systematic improvements in the MJO and QBO and in the teleconnection patterns associated with the PDO and ENSO. Some gains arise from better process representation, while others (e.g., the QBO) from higher resolution that allows for a greater range of interactions. Our results demonstrate that the incremental development processes in multiple climate model groups lead to more realistic simulations over time.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Leveraging national laboratories to increase Black representation in STEM: recommendations within the Department of Energy

Increasing diversity in STEM disciplines has been a goal at scientific institutions for many decades. Black representation in STEM, however, has remained critically low at all levels (high school, undergraduate, graduate, and professional) for over 40 years, highlighting the need for innovative strategies that promote and retain Black students and professionals in STEM. We refocus efforts on increasing Black representation in STEM by promoting early exposure and continued engagement while leveraging national laboratories—an underutilized resource with immense potential to centralize diversity and inclusion efforts nationally.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2024

ELM-Wet: Inclusion of a wetland landunit with sub-grid representation of eco-hydrological patches and hydrological forcing in E3SM Land Model (ELM)

Wetlands emit the most biogenic methane (CH4) and present the greatest uncertainty in the global CH4 budget. Modeling these emissions is challenging due to the temporal and spatial variability in wetland structure and CH4 flux rates, along with complex interactions among hydrological, ecological, meteorological, and microbial processes that govern CH4 dynamics. To address these issues, we aim to enhance the accuracy of wetland representation in the U.S. Department of Energy’s Exascale Earth System Model (E3SM) Land Model, ELM. This effort led to the development of ELM-Wet, which incorporates a dedicated wetland landunit with subgrid representation of eco-hydrological patch types. We implement wetland-specific hydrology by imposing site-specific constraints on surface water levels, thereby allowing different patches to sustain varying depths of inundation. Additionally, we refined the calculation of aerenchyma transport diffusivity based on observed conductance across different vegetation types. We validated these enhancements through site-specific simulations of a coastal freshwater wetland, the Salvador WMA Freshwater Marsh (Ameriflux, site ID US-LA2), located in the coast of Louisiana (29.85N,90.29W) at an elevation of 0 m. The site was simulated with ELM-Wet and the default version ELMv1. We use Bayesian Optimization to parameterize CO2 and CH4 fluxes. Eddy covariance observations of CO2 and CH4 fluxes from 2012-2013 were used to train the model and data from 2021 were used for validation. In this repository, we include all the output of all the simulations performed using different versions of ELMv1 and ELM-Wet, all input required to the model, and the Matlab scripts we used to processed the output data. ELM-Wet_flmd.csv includes a detailed description of the datasets files.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Concentric Spherical GNN for 3D Representation Learning

Learning 3D representations that generalize well to arbitrarily oriented inputs is a challenge of practical importance in applications varying from computer vision to physics and chemistry. We propose a novel multi-resolution convolutional architecture for learning over concentric spherical feature maps, of which the single sphere representation is a special case. Our hierarchical architecture is based on alternatively learning to incorporate both intra-sphere and inter-sphere information. We show the applicability of our method for two different types of 3D inputs, mesh objects, which can be regularly sampled, and point clouds, which are irregularly distributed. We also propose an efficient mapping of point clouds to concentric spherical images, thereby bridging spherical convolutions on grids with general point clouds. We demonstrate the effectiveness of our approach in improving state-of-the-art performance on 3D classification tasks with rotated data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

An examination of different representations of the electrostatic field and potential of an intense relativistic electron beam in a cylindrical conductor and their implications on foil focusing of an elliptical beam

Knowledge of the electrostatic field and potential of an intense relativistic electron beam (IREB) in a cylindrical conductor is essential for understanding the transport of IREBs including the interaction of a grounded conducting foil with the beam. Different representations of the electrostatic field or potential exist in the literature. The different representations are compared and an extension to foil focusing of elliptical beams is discussed.

43 PARTICLE ACCELERATORS↗
DOE OSTI2023

Improving the representation of isopycnal mixing in E3SM (Final Report)

The DOE-supported Energy Exascale Earth System Model (E3SM) is a major attempt by the Department of Energy to develop a new Earth System Model using unstructured grids, which allow for the model to concentrate resolution where it is most needed and avoid numerical artifacts experienced by previous Earth System Models (ESMs) using regular grids in which the lateral resolution becomes extremely fine as the resolution shrinks at the poles. This effort required rewriting many of the algorithms previously developed for regular grids. One algorithm that was problematic in the first version of E3SM was the representation of mixing due to turbulent ocean eddies with scales smaller than the model grid. These mesoscale eddies are the primary way by which tracers are stirred along density surfaces in the ocean interior. In particular, the process of isopycnal mixing exchanges fresher, more oxygenated waters from polar regions with saltier, nutrient-rich and oxygen poor waters from tropical regions. Previous work in our group has shown that this mechanism is vitally important for bringing oxygen into poorly ventilated tropical regions (Gnanadesikan et al., 2012, 2013) and plays a significant role in determining the uptake of anthropogenic carbon dioxide (Gnanadesikan et al., 2015). However, in E3SMv1 this process was turned off, as turning it on caused the model to become unstable. Additionally, the rate of isopycnal mixing is directly proportional to a mixing coefficient $A_{Redi}$ whose value varies from less than 400 m 2 /s to 2000 m 2 /s across contemporary climate models. The main thrusts of this proposal therefore were to: 1.) Improve the numerics of isopycnal mixing in E3SM; 2.) Identify ways in which climate models are sensitive to the isopycnal mixing coefficient; and 3.) Develop better representations of the isopycnal mixing coefficient.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Understanding the Physics Representation of Deep Learning Models in Environmental Applications

Deep learning (DL) models have been popular in earth and environmental modeling and analysis, which exhibit huge potential in capturing and reconstructing the non-linearity of relevant environmental processes. They are extensively used as analytical tools or emulators for multiple domains (atmosphere, land surface, ocean, and biogeochemistry). Despite their success, their internal working mechanism remains largely unknown. Such a lack of knowledge hinders the identification of physically consistent models that are fully adaptive to non-stationary climate, as well as the development of physics-informed machine learning such as physics-informed neural network (PINN). To establish preliminary knowledge and framework of such physics representation evaluation, this project focuses on an improved understanding of DL models in the environmental applications. DL models are increasingly applied to environmental modeling and prediction. However, they have been evaluated mostly from a performance perspective, and there is a gap in understanding how they represent the known physics internally. Such knowledge is especially critical when applying DL models under climate change conditions, where new inputs are likely outside the ranges of the training datasets. In this project, we reveal how the known physical processes are represented within DL models from both statistical and mechanistic perspectives. Leveraging the traditional model evaluations that focus more on the accuracies of predictions, we establish a framework that examines both the accuracy and physics representation of DL models. This analysis framework can identify DL models that make the correct predictions based on correct physics, thus enhancing the existing explainable artificial intelligence (explainable-AI) portfolio. It lays a foundation for developing novel metrics to evaluate the emerging DL models in environmental applications. This knowledge also informs the development of physics-informed DL models by revealing the direct connections between the known physical processes and specific model components or structures.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Towards an improved representation of carbonaceous aerosols over the Indian monsoon region in a regional climate model: RegCM

Mitigation of carbonaceous aerosol emissions is expected to provide climate and health co-benefits. The accurate representation of carbonaceous aerosols in climate models is critical for reducing uncertainties in their climate feedback. In this regard, emission fluxes and aerosol life cycle processes are the two primary sources of uncertainties. Here, we demonstrate that the incorporation of a dynamic ageing scheme and emission estimates that are updated for the local sources improves the representation of carbonaceous aerosols over the Indian monsoon region in a regional climate model, RegCM, compared with its default configuration. The respective mean black carbon (BC) and organic carbon (OC) surface concentrations in 2010 are estimated to be 4.25 and 10.35 µg m -3 over the Indo-Gangetic Plain (IGP) in the augmented model. The BC column burden over the polluted IGP is found to be 2.47 mg m -2 , 69.95 % higher than in the default model configuration and much closer to available observations. The anthropogenic aerosol optical depth (AOD) increases by more than 19 % over the IGP due to the model enhancement, also leading to a better agreement with observed AOD. The respective top-of-the-atmosphere, surface, and atmospheric anthropogenic aerosol short-wave radiative forcing are estimated at -0.3, -9.3, and 9.0 W m -2 over the IGP and -0.89, -5.33, and 4.44 W m -2 over Peninsular India (PI). Our results suggest that the combined effect of two modifications leads to maximum improvements in the model performance in regions where emissions play a dominant role.

54 ENVIRONMENTAL SCIENCES↗