SEARCH · Engineering Papers

Results for “distributed algorithms”

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 505 records · Page 28

All sources NASA NTRS DOE OSTI

DOE OSTI2022

SMALE: Enhancing Scalability of Machine Learning Algorithms on Extreme-Scale Computing Platforms

Deployment and execution of machine learning tasks on extreme-scale computing platforms face several significant technical challenges: 1) High computing cost incurred by dense networks – The computing workload of deep networks with densely-connected topology increases rapidly with the network size, imposing a non-scalable computing model of extreme-scale computing platforms; 2) Non-optimized workload distribution – Many advanced deep learning algorithms, e.g., sparsification and irregular net-work topology, produce very unbalanced workload distribution on extreme-scale computing platforms. The computation efficiency is greatly hindered by the incurred data and computation redundancies as well as long tails of the node with extensive workload; 3) Constraints in data movement and I/O bottle-neck – Inter-node data movement in extreme-scale computing platforms are associated with high energy and latency costs, and subject to the constraints of I/O bandwidth; and 4) Generalization of algorithm realization and acceleration on computing platforms – The large varieties of machine learning algorithms and structures of extreme-scale computing platforms make the derivation of a generalized algorithm realization and acceleration method very challenging, which, however, is the requirement by domain scientists and interested users. We call the above challenges Smale’s Problems in Machine Learning and Understanding for High-Performance Computing Scientific Discovery. The objective of our three-year research project is to develop a holistic innovation set at structure, assembly, and acceleration layers of machine learning algorithms to address the above challenges in algorithm deployment and execution. Three tasks are particularly performed, including: At the algorithm structure level, we investigate the techniques that can structurally sparsify on the topology of deep networks for computing workload reduction. We also study clustering and pruning techniques that can optimize the workload distributions over the extreme-scale computing platforms; At the algorithm assembly level, we derive a unified learning framework for unsupervised transfer learning and dynamic growing capabilities. Novel training methods are also exploited to enhance the training efficiency of the proposed framework; At the algorithm acceleration level, we will develop a series of techniques that can accelerate the computation of sparse matrix operations, which are one of the core executions in deep learning and optimize memory access of the concerned platforms. Our proposed techniques attack the fundamental problems in machine learning algorithms running on extreme-scale computing platforms by vertically integrating the solutions at three closely entangled layers, paving the long-term scaling path of machine learning applications under DOE context. Three tasks corresponding to the above respective research orientations are performed during the three-year project period with our collaborators at ORNL. The outcome of the proposed project is anticipated to form a holistic solution set of novel algorithms and network topologies, efficient training techniques, and fast acceleration methods to promote the computing scalability of the machine learning applications of particular interest to DOE.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2018

Application of chord length distributions and principal component analysis for quantification and representation of diverse polycrystalline microstructures

Highlights: • Microstructure quantification based on directional chord length distribution and principal component analysis is presented • A boundary-based algorithm of calculations of chord length distributions is introduced • Utility of the quantification framework is shown in three case studies on synthetic, experimental, and simulation datasets • Angularly resolved chord length distribution captures essential details of grain size and shape in diverse microstructures • Low-dimensional representations of directional chord length distributions are obtained using principal component analysis - Abstract: Quantification of mesoscale microstructures of polycrystalline materials is important for a range of practical tasks of materials design and development. The current protocols of quantifying grain size and morphology often rely on microstructure metrics (e.g., mean grain diameter) that overlook important details of the mesostructure. In this work, we present a quantification framework based on directionally resolved chord length distribution and principal component analysis as a means of extracting additional information from 2-D microstructural maps. Towards this end, we first present in detail a method for calculating chord length distribution based on boundary segments available in modern digital datasets (e.g., from microscopy post-processing) and their low-rank representations by principal component analysis. The utility of the proposed framework for capturing grain size, morphology, and their anisotropy for efficient visualization, representation, and specification of polycrystalline microstructures is then demonstrated in case studies on datasets from synthetic generation, experiments (on Ni-base superalloys), and simulations (on steel during recrystallization).

36 MATERIALS SCIENCE↗

DOE OSTI2015

Identifying Important Ions and Positions in Mass Spectrometry Imaging Data Using CUR Matrix Decompositions

Mass spectrometry imaging enables label-free, high-resolution spatial mapping of the chemical composition of complex, biological samples. Typical experiments require selecting ions and/or positions from the images: ions for fragmentation studies to identify keystone compounds and positions for follow up validation measurements using microdissection or other orthogonal techniques. Unfortunately, with modern imaging machines, these must be selected from an overwhelming amount of raw data. Existing techniques to reduce the volume of data, the most popular of which are principle component analysis and non-negative matrix factorization, have the disadvantage that they return difficult-to-interpret linear combinations of actual data elements. In this work, we show that CX and CUR matrix decompositions can be used directly to address this selection need. CX and CUR matrix decompositions use empirical statistical leverage scores of the input data to provide provably good low-rank approximations of the measured data that are expressed in terms of actual ions and actual positions, as opposed to difficult-to-interpret eigenions and eigenpositions. We show that this leads to effective prioritization of information for both ions and positions. In particular, important ions can be found either by using the leverage scores as a ranking function and using a deterministic greedy selection algorithm or by using the leverage scores as an importance sampling distribution and using a random sampling algorithm; however, selection of important positions from the original matrix performed significantly better when they were chosen with the random sampling algorithm. Also, we show that 20 ions or 40 locations can be used to reconstruct the original matrix to a tolerance of 17% error for a widely studied image of brain lipids; and we provide a scalable implementation of this method that is applicable for analysis of the raw data where there are often more than a million rows and/or columns, which is larger than SVD-based low-rank approximation methods can handle. Lastly, these results introduce the concept of CX/CUR matrix factorizations to mass spectrometry imaging, describing their utility and illustrating principled algorithmic approaches to deal with the overwhelming amount of data generated by modern mass spectrometry imaging.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2012

A Hybrid Demand Response Simulator Version 1.0

A hybrid demand response simulator is developed to test different control algorithms for centralized and distributed demand response (DR) programs in a small distribution power grid. The HDRS is designed to model a wide variety of DR services such as peak having, load shifting, arbitrage, spinning reserves, load following, regulation, emergency load shedding, etc. The HDRS does not model the dynamic behaviors of the loads, rather, it simulates the load scheduling and dispatch process. The load models include TCAs (water heaters, air conditioners, refrigerators, freezers, etc) and non-TCAs (lighting, washer, dishwasher, etc.) The ambient temperature changes, thermal resistance, capacitance, and the unit control logics can be modeled for TCA loads. The use patterns of the non-TCA can be modeled by probability of use and probabilistic durations. Some of the communication network characteristics, such as delays and errors, can also be modeled. Most importantly, because the simulator is modular and greatly simplified the thermal models for TCA loads, it is very easy and fast to be used to test and validate different control algorithms in a simulated environment.

Lu, Ning↗

DOE OSTI2016

Distribution system model calibration with big data from AMI and PV inverters

Efficient management and coordination of distributed energy resources with advanced automation schemes requires accurate distribution system modeling and monitoring. Big data from smart meters and photovoltaic (PV) micro-inverters can be leveraged to calibrate existing utility models. This paper presents computationally efficient distribution system parameter estimation algorithms to improve the accuracy of existing utility feeder radial secondary circuit model parameters. The method is demonstrated using a real utility feeder model with advanced metering infrastructure (AMI) and PV micro-inverters, along with alternative parameter estimation approaches that can be used to improve secondary circuit models when limited measurement data is available. Lastly, the parameter estimation accuracy is demonstrated for both a three-phase test circuit with typical secondary circuit topologies and single-phase secondary circuits in a real mixed-phase test system.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2016

A whole-path importance-sampling scheme for Feynman path integral calculations of absolute partition functions and free energies

Using Feynman path integrals, a molecular partition function can be written as a double integral with the inner integral involving all closed paths centered at a given molecular configuration, and the outer integral involving all possible molecular configurations. In previous work employing Monte Carlo methods to evaluate such partition functions, we presented schemes for importance sampling and stratification in the molecular configurations that constitute the path centroids, but we relied on free-particle paths for sampling the path integrals. At low temperatures, the path sampling is expensive because the paths can travel far from the centroid configuration. In this paper, we now present a scheme for importance sampling of whole Feynman paths based on harmonic information from an instantaneous normal mode calculation at the centroid configuration, which we refer to as harmonically guided whole-path importance sampling (WPIS). We obtain paths conforming to our chosen importance function by rejection sampling from a distribution of free-particle paths. Sample calculations on CH 4 demonstrate that at a temperature of 200 K, about 99.9% of the free-particle paths can be rejected without integration, and at 300 K, about 98% can be rejected. Finally, we also show that it is typically possible to reduce the overhead associated with the WPIS scheme by sampling the paths using a significantly lower-order path discretization than that which is needed to converge the partition function.

74 ATOMIC AND MOLECULAR PHYSICS↗

DOE OSTI2025

Resilient Operation of Networked Community Microgrids with High Solar Penetration

This project, funded by the US Department of Energy’s Solar Energy Technologies Office (SETO), focused on the operation of microgrids as a coordinated network. The primary objective, which was successfully achieved, was to develop both control strategies and hardware solutions to support the resilient and efficient operation of networked microgrids with high solar penetration. The work was structured around the following four main tasks: • Development of distributed and scalable optimization algorithms for AC-coupled networked microgrids. • Design and implementation of a novel DC interconnection hardware to enable precise power exchange between microgrids. • Laboratory operational validation of the developed technologies using 480 V testbeds and commercially available hardware. • Field operational validation of the complete solution in Adjuntas, Puerto Rico, interconnecting two kW-scale, split-phase microgrids of Casa Pueblo’s microgrids. This project addressed multiple technical challenges across the domains of optimization, control, hardware interconnection, and protection. One of its key contributions was delivering tangible, real-world solutions for networking microgrids. In contrast to purely theoretical or simulation-based work, this project included full-scale hardware operational validation both in the lab and in the field. The work conducted as part of this project—in collaboration with the University of Puerto Rico; the University of Tennessee, Knoxville; the University of Central Florida; and Casa Pueblo—has advanced the state of the art in networked microgrids. Key contributions include the development of distributed control strategies, practical solutions for real-world implementation challenges, and the introduction of a novel DC interlink approach for microgrid interconnection. The project featured both laboratory and field validation using commercial off-the-shelf components. The field deployment successfully validated that a group of microgrids can operate in a coordinated manner, enabling precise power flow between systems and mutual support during extreme events. This project resulted in 15 journal publications and 15 conference papers; 5 graduate students and 15 undergraduate students were supported. The codes of distributed optimization and forecasting were made open-source through OSTI.gov for distributed optimization and forecasting. All the publications are available in the ORNL-hosted project landing page. The DC interlink with state-of-charge balancing control was operationally validated in Adjuntas by interconnecting two real-world, 240 V split-phase microgrids. To the best knowledge of the team, this represents the first operational validation of AC microgrids interconnected via DC-interlinks. As a culmination of this project, a follow-on grant was awarded to support the technology transfer of the distributed optimization framework to a commercial microgrid controller, Stellar Edge, developed by the California-based company New Sun Road.

14 SOLAR ENERGY↗

DOE OSTI2025

Distributed Inertia Management Under Communication Constraints

This paper presents a framework for distributed inertia management based on consensus algorithm. We propose a methodology to achieve optimal operation of inertia sources such as distributed energy resources (DERs) and synchronous generators in real time. Additionally, we analyze the algorithm under communication constraints and evaluate its robustness under scenarios involving communication time delays and packet losses. The proposed approach is validated via simulations on a 4-node system test feeder, demonstrating its effectiveness and resilience.

Yadav, Ajay [ORNL] (ORCID:000000016111881X)↗

DOE OSTI2021

Operating a commercial building HVAC load as a virtual battery through airflow control

Virtual battery (VB) is an innovative method to model flexibility of building loads and effectively coordinate them with other resources at a system level. Unlike a real battery with a dedicated power conversion system for charging control, methods are required for operating building loads to deviate from the baseline to respond to grid signals. This paper presents a VB control for a commercial heating, ventilation, and air conditioning (HVAC) system to follow the desired power consumption in real-time by adjusting zonal airflow rates. The proposed method consists of two parts. At the system level, a mixed feedforward and feedback control is used to estimate the desired total airflow rate. At the zone level, two priority-based algorithms are then proposed to distribute the total airflow rate to individual zones. In particular, a zonal airflow limit estimation method is proposed using machine-learning techniques, in contrast to physics-based thermal models in existing studies, to more accurately capture zonal thermal dynamics and improve temperature control performance. An office building on the Pacific Northwest National Laboratory campus is implemented in EnergyPlus, and used to illustrate and validate the proposed control.

Wang, Jiyu↗

DOE OSTI2017

A Review of Power Distribution Test Feeders in the United States and the Need for Synthetic Representative Networks

Under the increasing penetration of distributed energy resources and new smart network technologies, distribution utilities face new challenges and opportunities to ensure reliable operations, manage service quality, and reduce operational and investment costs. Simultaneously, the research community is developing algorithms for advanced controls and distribution automation that can help to address some of these challenges. However, there is a shortage of realistic test systems that are publically available for development, testing, and evaluation of such new algorithms. Concerns around revealing critical infrastructure details and customer privacy have severely limited the number of actual networks published and that are available for testing. In recent decades, several distribution test feeders and US-featured representative networks have been published, but the scale, complexity, and control data vary widely. This paper presents a first-of-a-kind structured literature review of published distribution test networks with a special emphasis on classifying their main characteristics and identifying the types of studies for which they have been used. As a result, this both aids researchers in choosing suitable test networks for their needs and highlights the opportunities and directions for further test system development. In particular, we highlight the need for building large-scale synthetic networks to overcome the identified drawbacks of current distribution test feeders.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2022

Hierarchical Optimal Power Flow with Improved Gradient Evaluation

Existing algorithms to solve alternating-current optimal power flow (AC-OPF) often exploit linear approximations to simplify system models and accelerate computations. In this paper, we improve a recent hierarchical OPF algorithm, which rested on primal-dual gradients evaluated in a linearized distribution power flow model. Specifically, we identify a risk of voltage violation arising from the model linearization, and propose a more accurate gradient evaluation method to eliminate that risk. We further develop a hierarchical primal-dual algorithm to solve OPF based on the proposed gradient evaluation method. Numerical results on IEEE networks show that our algorithm can enhance voltage safety with satisfactory computational efficiency.

distributed algorithm↗

DOE OSTI2025

Parallel sorting algorithm classification: is manual instrumentation necessary?

Understanding parallel algorithms is crucial for accelerating scientific simulations on complex, distributed memory, high-performance computers. Modern algorithm classification approaches learn semantics directly from source code to differentiate between algorithms, however, accessing source code is not always possible. We can learn about parallel algorithms from observing their performance, as programs running the same algorithms and using the same hardware should exhibit similar performance characteristics. We present an approach to learn algorithm classes from parallel performance data directly in order to classify algorithms without access to the source code. We extend previous work to enable classifying parallel sorting algorithms using automatic instrumentation instead of requiring manual region annotations in the source code. In this work, we design and demonstrate a study for classification of parallel sorting algorithms using parallel performance data collected from automatic instrumentation, and evaluate the performance of our new methodology on classification. We leverage Caliper to collect the performance data, Thicket for our exploratory data analysis (EDA), and PyTorch and Scikit-learn to evaluate the effectiveness of random forests, support vector machines (SVMs), decision trees, neural networks, and logistic regressions on parallel performance data. Additionally, we study noise in parallel performance data, whether the removal of noise and pre-processing of the data is necessary to accurately classify parallel sorting algorithms, and determine the effectiveness of features created from performance data. In conclusion, we demonstrate classification accuracy for these five different models of up to 97.7% across four different parallel algorithm classes.

Algorithm Classification↗

DOE OSTI2013

Optimal field-splitting algorithm in intensity-modulated radiotherapy: Evaluations using head-and-neck and female pelvic IMRT cases

To develop an optimal field-splitting algorithm of minimal complexity and verify the algorithm using head-and-neck (H and N) and female pelvic intensity-modulated radiotherapy (IMRT) cases. An optimal field-splitting algorithm was developed in which a large intensity map (IM) was split into multiple sub-IMs (≥2). The algorithm reduced the total complexity by minimizing the monitor units (MU) delivered and segment number of each sub-IM. The algorithm was verified through comparison studies with the algorithm as used in a commercial treatment planning system. Seven IMRT, H and N, and female pelvic cancer cases (54 IMs) were analyzed by MU, segment numbers, and dose distributions. The optimal field-splitting algorithm was found to reduce both total MU and the total number of segments. We found on average a 7.9 ± 11.8% and 9.6 ± 18.2% reduction in MU and segment numbers for H and N IMRT cases with an 11.9 ± 17.4% and 11.1 ± 13.7% reduction for female pelvic cases. The overall percent (absolute) reduction in the numbers of MU and segments were found to be on average −9.7 ± 14.6% (−15 ± 25 MU) and −10.3 ± 16.3% (−3 ± 5), respectively. In addition, all dose distributions from the optimal field-splitting method showed improved dose distributions. The optimal field-splitting algorithm shows considerable improvements in both total MU and total segment number. The algorithm is expected to be beneficial for the radiotherapy treatment of large-field IMRT.

61 RADIATION PROTECTION AND DOSIMETRY↗

DOE OSTI2015

Electron dose distributions caused by the contact-type metallic eye shield: Studies using Monte Carlo and pencil beam algorithms

A metallic contact eye shield has sometimes been used for eyelid treatment, but dose distribution has never been reported for a patient case. This study aimed to show the shield-incorporated CT-based dose distribution using the Pinnacle system and Monte Carlo (MC) calculation for 3 patient cases. For the artifact-free CT scan, an acrylic shield machined as the same size as that of the tungsten shield was used. For the MC calculation, BEAMnrc and DOSXYZnrc were used for the 6-MeV electron beam of the Varian 21EX, in which information for the tungsten, stainless steel, and aluminum material for the eye shield was used. The same plan was generated on the Pinnacle system and both were compared. The use of the acrylic shield produced clear CT images, enabling delineation of the regions of interest, and yielded CT-based dose calculation for the metallic shield. Both the MC and the Pinnacle systems showed a similar dose distribution downstream of the eye shield, reflecting the blocking effect of the metallic eye shield. The major difference between the MC and the Pinnacle results was the target eyelid dose upstream of the shield such that the Pinnacle system underestimated the dose by 19 to 28% and 11 to 18% for the maximum and the mean doses, respectively. The pattern of dose difference between the MC and the Pinnacle systems was similar to that in the previous phantom study. In conclusion, the metallic eye shield was successfully incorporated into the CT-based planning, and the accurate dose calculation requires MC simulation.

61 RADIATION PROTECTION AND DOSIMETRY↗

DOE OSTI2017

Joint Bayesian Estimation of Quasar Continua and the Lyα Forest Flux Probability Distribution Function

We present a new Bayesian algorithm making use of Markov Chain Monte Carlo sampling that allows us to simultaneously estimate the unknown continuum level of each quasar in an ensemble of high-resolution spectra, as well as their common probability distribution function (PDF) for the transmitted Lyα forest flux. This fully automated PDF regulated continuum fitting method models the unknown quasar continuum with a linear principal component analysis (PCA) basis, with the PCA coefficients treated as nuisance parameters. The method allows one to estimate parameters governing the thermal state of the intergalactic medium (IGM), such as the slope of the temperature–density relation γ−1, while marginalizing out continuum uncertainties in a fully Bayesian way. Using realistic mock quasar spectra created from a simplified semi-numerical model of the IGM, we show that this method recovers the underlying quasar continua to a precision of ≃7% and ≃10% at z = 3 and z = 5, respectively. Given the number of principal component spectra, this is comparable to the underlying accuracy of the PCA model itself. Most importantly, we show that we can achieve a nearly unbiased estimate of the slope γ−1 of the IGM temperature–density relation with a precision of ±8.6% at z = 3 and ±6.1% at z = 5, for an ensemble of ten mock high-resolution quasar spectra. Applying this method to real quasar spectra and comparing to a more realistic IGM model from hydrodynamical simulations would enable precise measurements of the thermal and cosmological parameters governing the IGM, albeit with somewhat larger uncertainties, given the increased flexibility of the model.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2021

HTR-1.3 solver: Predicting electrified combustion using the hypersonic task-based research solver

Here this manuscript presents an updated open-source version of the Hypersonics Task-based Research (HTR) solver. The solver, whose main features are presented in Di Renzo et al. (2020) and Di Renzo & Pirozzoli (2021), is designed for direct numerical simulation of reacting flows at high Reynolds numbers. This new version extends the applications of the HTR solver to turbulent combustion in the presence of external electric fields. In particular, a new distributed Poisson solver compatible with heterogeneous architectures has been incorporated in the algorithm to compute the electric potential distribution in bi-periodic configurations. The drift fluxes of the electrically charged species are now included in the transport equations using a targeted essentially non-oscillatory scheme. A verification of these new features of the solver is provided using one-dimensional burner stabilized flames, whereas a three dimensional turbulent flame is utilized to discuss the scalability of the proposed numerical tool.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2019

Rapid simulation of X-ray scatter measurements for threat detection via GPU-based ray-tracing

Scatter-based X-ray imaging has increased in prominence in areas ranging from cancer diagnostics to threat detection in aviation security, due largely to the development of new algorithms and components over the last decade. However, system design and algorithm development are often hindered by an inability to generate a sucient amount of accurate data. To further the development of scatter-based systems, we created a rapid X-ray scatter simulation tool built with a GPU-centric, parallel ray-tracing framework (NVIDIA OptiX). This tool models a full range of X-ray imaging components and describes 3D objects formed with heterogeneous media using polygon meshes. The scatter simulation algorithm we used is similar to the previously-described hybrid approach; however, instead of employing Monte Carlo techniques, we developed a purely analytical algorithm to sample the distribution of single-scatter events throughout the region of interest. The contribution of scattered photons to the measured signal is then calculated using the first-Born approximation. Here, the accuracy of the implemented pipeline has been validated by comparing simulated against experimental data obtained with a laboratory X-ray scattering system and phantoms of varied materials. Using a single desktop computer with an NVIDIA GTX 770 GPU, we show that the scatter signal generated by an incident fan beam and recorded by a 125 125 element detector array can be simulated in on the order of a few to tens of minutes (depending on the object extent and complexity). As a point of comparison, simulations performed via CPU-based Monte Carlo tools, such as GEANT4 (i.e., the gold standard), can take up to tens of hours to achieve comparable results.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2021

Understanding kinetics of defect annihilation in chemoepitaxy-directed self-assembly.

Directed self-assembly (DSA) of block copolymers (BCP) has attracted considerable interest from the semiconductor industry because it can achieve semiconductor-relevant structures with a relatively simple process and low cost. However, the self-assembling structures can become kinetically trapped into defective states, which greatly impedes the implementation of DSA in high-volume manufacturing. Understanding the kinetics of defect annihilation is crucial to optimizing the process and eventually eliminating defects in DSA. Such kinetic experiments, however, are not commonly available in academic laboratories. To address this challenge, we perform a kinetic study of chemoepitaxy DSA in a 300 mm wafer fab, where the complete defectivity information at various annealing conditions can be readily captured. Through extensive statistical analysis, we reveal the statistical model of defect annihilation in DSA for the first time. The annihilation kinetics can be well described by a power law model, indicating that all dislocations can be removed by sufficiently long annealing time. We further develop image analysis algorithms to analyze the distribution of dislocation size and configurations and discover that the distribution stays relatively constant over time. The defect distribution is determined by the role of the guiding stripe, which is found to stabilize the defects. Although this study is based on polystyrene-b-poly(methyl methacrylate) (PS-b-PMMA), we anticipate that these findings can be readily applied to other BCP platforms as well.

block copolymer↗