Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “distributed computation”

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 1,351 records · Page 75
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Multiperiod Distribution System Restoration With Routing Repair Crews, Mobile Electric Vehicles, and Soft-Open-Point Networked Microgrids

This paper proposes a distribution system restoration model which is in response to multiple outages caused by natural disasters. The proposed restoration model includes the coordination of routing repair crews (RRCs), mobile batterycarried vehicles (MBCVs), and networked microgrids (NMGs) formed by soft open points (SOPs). The travel and repair time constraints are modeled for each RRC; travel path and charging strategy are modeled for each MBCV; and the network reconfiguration is developed considering the optimal operation of SOP-based NMGs. Furthermore, the proposed model is presented as a mixedinteger linear program which is solved by an auxiliary induce function based algorithm to reduce the computational complexity. The modified IEEE 33-bus and 69-bus distribution systems are tested with multiple outages. The presented results demonstrate the effectiveness of the proposed model

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

An adaptive Hessian approximated stochastic gradient MCMC method

Bayesian approaches have been successfully integrated into training deep neural networks. One popular family is stochastic gradient Markov chain Monte Carlo methods (SG-MCMC), which have gained increasing interest due to their ability to handle large datasets and the potential to avoid overfitting. Although standard SG-MCMC methods have shown great performance in a variety of problems, they may be inefficient when the random variables in the target posterior densities have scale differences or are highly correlated. Here, we present an adaptive Hessian approximated stochastic gradient MCMC method to incorporate local geometric information while sampling from the posterior. The idea is to apply stochastic approximation (SA) to sequentially update a preconditioning matrix at each iteration. The preconditioner possesses second-order information and can guide the random walk of a sampler efficiently. Instead of computing and saving the full Hessian of the log posterior, we use limited memory of the samples and their stochastic gradients to approximate the inverse Hessian-vector multiplication in the updating formula. Moreover, by smoothly optimizing the preconditioning matrix via SA, our proposed algorithm can asymptotically converge to the target distribution with a controllable bias under mild conditions. To reduce the training and testing computational burden, we adopt a magnitude-based weight pruning method to enforce the sparsity of the network. Our method is user-friendly and demonstrates better learning results compared to standard SG-MCMC updating rules. The approximation of inverse Hessian alleviates storage and computational complexities for large dimensional models. Numerical experiments are performed on several problems, including sampling from 2D correlated distribution, synthetic regression problems, and learning the numerical solutions of heterogeneous elliptic PDE. The numerical results demonstrate great improvement in both the convergence rate and accuracy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Distributed deep learning training using silicon photonic switched architectures

The scaling trends of deep learning models and distributed training workloads are challenging network capacities in today’s datacenters and high-performance computing (HPC) systems. We propose a system architecture that leverages silicon photonic (SiP) switch-enabled server regrouping using bandwidth steering to tackle the challenges and accelerate distributed deep learning training. In addition, our proposed system architecture utilizes a highly integrated operating system-based SiP switch control scheme to reduce implementation complexity. To demonstrate the feasibility of our proposal, we built an experimental testbed with a SiP switch-enabled reconfigurable fat tree topology and evaluated the network performance of distributed ring all-reduce and parameter server workloads. The experimental results show up to 3.6× improvements over the static non-reconfigurable fat tree. Our large-scale simulation results show that server regrouping can deliver up to 2.3× flow throughput improvement for a 2× tapered fat tree and a further 11% improvement when higher-layer bandwidth steering is employed. The collective results show the potential of integrating SiP switches into datacenters and HPC systems to accelerate distributed deep learning training.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

Coupled Electro-Thermal Analysis of Permanent Magnet Synchronous Motor for Electric Vehicles

Permanent magnet synchronous motors (PMSM) are extensively used in electric vehicles. However, high internal heat generation and inefficient heat dissipation often limit the operational reliability, and longevity of the PMSM. Therefore, proper quantification of heat generation in electric motor and advanced embedded motor cooling techniques remain topics of immense interest. In order to accurately predict electro-magnetic performance, i.e., efficiency, component-wise heat losses and the corresponding temperature distribution of a jacket cooled machine, this paper presents a two-way iteratively coupled electro-thermal modeling framework. Finite element based software Motor-CAD has been utilized for electro-magnetic performance calculation of BMW i3 PMSM. Finite volume based computational fluid dynamics/heat transfer (CFD/HT) software ANSYS® FLUENT® has been employed to simulate the temperature distribution of the PMSM, using the electro-magnetic losses as heat input. Computed heat losses, stator, winding, rotor, and magnet temperatures are utilized as coupling parameters between the electro-magnetic and thermal models. A conventional one-way coupling algorithm has also been developed and compared to the newly proposed two-way coupling algorithm. Numerical results confirm that at high current density, one-way coupling algorithm significantly over-predicts the motor temperature compared to the two-way algorithm. A comprehensive analysis has been carried out to characterize the influences of current density, speed, and forced convection heat transfer coefficient on the heat losses, overall efficiency, and maximum temperature of the PMSM. Lastly, an efficiency map has been interpreted from the coupled electro-magnetic simulation.

42 ENGINEERING↗
DOE OSTI2023

Software-Hardware Co-design of Heterogeneous SmartNIC System for Recommendation Models Inference and Training

Deep Learning Recommendation Models (DLRMs) are critical applications in various domains and have evolved as one of the single largest machine learning applications. Trillions of DLRM parameters exceed the on-chip memory capacity of GPUs. Large-scale multi-node systems are required for distributed DLRM inference and training, which suffer from the all-to-all communication bottleneck, mainly limiting the scalability of ever-growing DLRMs. In recent years, SmartNICs have evolved with coupled computation and communication capabilities providing opportunities for a powerful heterogeneous device in the system. However, there isn't such a distributed system that fully leverages the abundant smartNIC resources that resolve the scalability issue of DLRMs. In this work, we proposed a software-hardware co-design of a heterogeneous smartNIC system that resolves the communication bottleneck of distributed DLRMs, mitigates the memory bandwidth pressure, and improves computation efficiency. We provide a set of smartNIC designs of cache systems (including local cache and remote cache) and smartNIC computation kernels which reduce data movement, relieve memory lookup intensity, and improve the GPU's computation efficiency. In addition, we propose a graph algorithm that improves the data locality of queries within batches which optimizes the overall system performance with higher data reuse. Our evaluation shows that our system achieves 2.1x latency speedup for inference and 1.6x throughput speedup for training.

Guo, Anqi↗
DOE OSTI2021

An Energy Service Interface for Distributed Energy Resources

Renewable energy resources, particularly wind and solar photovoltaic, are becoming significant contributors to electric power generation. These re-sources will contribute towards achieving sustainable electric power systems. However, renewable resources will dramatically increase the demand for flexible power system operations. This paper proposes an energy service interface that will allow aggregated distributed energy resources, such as residential loads and inverter-based systems, to participate in NERC-defined smart energy reliability services. Such cyber-physical systems will increase system flexibility by ensuring match between energy supply and energy demand.Aggregation and coordinated dispatch of millions of distributed energy resources will require development of large-scale computing networks. Several smart grid interface-enabling technologies, including IEEE 2030.5, Common Smart Inverter Profile, SunSpec Modbus, and CTA 2045, are discussed. Residential loads are categorized by their static and dynamic energy characteristics to identify services in which they can participate. The business model for the energy services interface as well as probabilistic modeling for resource estimation are highlighted as future considerations.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

Computational design of microarchitected porous electrodes for redox flow batteries

Porous electrodes are used as the core reactive component across electrochemical technologies. In flowing systems, controlling the fluid distribution, species transport, and reactive environment is critical to attaining high performance. However, conventional electrode materials like felts and papers provide few opportunities for precise engineering of the electrode and its microstructure. To address these limitations, architected electrodes composed of unit cells with spatially varying geometry determined via computational optimization are proposed. Resolved simulation is employed to develop a homogenized description of the constituent unit cells. These effective properties serve as inputs to a continuum model for the electrode when used in the negative half-cell of a vanadium redox flow battery. Porosity distributions minimizing power loss are then determined via computational design optimization to generate architected porosity electrodes. The architected electrodes are compared to bulk, uniform porosity electrodes and found to lead to increased power efficiency across operating flow rates and currents. The design methodology is further used to generate a scaled-up electrode with comparable power efficiency to the bench-scale systems. Finally, the variable porosity architecture and computational design methodology presented here thus offers a novel pathway for automatically generating spatially engineered electrode structures with improved power performance.

25 ENERGY STORAGE↗
DOE OSTI2022

PV Hosting Capacity Estimation: Experiences with Scalable Framework; Preprint

Hosting capacity is an indication of the amount of photovoltaics (PV) can be hosted in a distribution system. This paper presents a framework for estimating distributed PV hosting capacity at scale. We first analyze the key challenges of performing relevant large scale simulation including computational and modeling challenges. Then, we develop two python-based software tools in order to conduct repeatable distribution analyses: Distribution Integration Solution Cost Options (DISCO) for configuring and analyzing simulations, and JADE for parallelizing jobs on HPC clusters. A case study of hosting capacity estimation for SMART-DS SFO 2000+ synthetic feeders is used to demonstrate the capability of the developed framework and tools. The framework and tools can help utilities assess the overall hosting capacity of their service territory, through which the overall upgrade cost can be better planned in order to integrate more PV in the future.

distributed energy resources↗
DOE OSTI2021

Distributed Optimization for Nonrigid Nano-Tomography

Resolution level and reconstruction quality in nano-computed tomography (nano-CT) are in part limited by the stability of microscopes, because the magnitude of mechanical vibrations during scanning becomes comparable to the imaging resolution, and the ability of the samples to resist radiation induced deformations during data acquisition. In such cases, there is no incentive in recovering the sample state at different time steps like in time-resolved reconstruction methods, but instead the goal is to retrieve a single reconstruction at the highest possible spatial resolution and without any imaging artifacts. Here we propose a distributed optimization solver for tomographic imaging of samples at the nanoscale. Our approach solves the tomography problem jointly with projection data alignment, nonrigid sample deformation correction, and regularization. Projection data consistency is regulated by dense optical flow estimated by Farneback's algorithm, leading to sharp sample reconstructions with less artifacts. Synthetic data tests show robustness of the method to Poisson and low-frequency background noise. We accelerated the solver on multi-GPU systems and validated the method on three nano-imaging experimental data sets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

On the computation of moments in the Super-Transition-Arrays model for radiative opacity calculations

In the Super-Transition-Array statistical method for the computation of radiative opacity of hot dense matter, the moments of the absorption or emission features involve partition functions with reduced degeneracies, occurring through the calculation of averages of products of subshell populations. Here, in the present work, we discuss several aspects of the computation of such peculiar partition functions, insisting on the precautions that must be taken in order to avoid numerical difficulties. In a previous work, we derived a formula for supershell partition functions, which takes the form of a functional of the distribution of energies within the supershell and allows for fast and accurate computations, truncating the number of terms in the expansion. The latter involves coefficients for which we obtained a recursion relation and an explicit formula. We show that such an expansion can be combined with the recurrence relation for shifted partition functions. We also propose, neglecting the effect of fine structure as a first step, a positive-definite formula for the Super-Transition-Array moments of any order, providing an insight into the asymmetry and sharpness of the latter. The corresponding formulas are free of alternating sums. Several ways to speed up the calculations are also presented.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2022

Data-Driven Probabilistic Voltage Risk Assessment of MiniWECC System with Uncertain PVs and Wind Generations using Realistic Data

Here, it is found from actual data that due to generation dispatch and uncertain renewable generations and loads with complicated correlations, inferring the probabilistic distributions for uncertain inputs is challenging. Many probabilistic power flow approaches have been developed in the literature but their validations using realistic systems and data are lacking. This paper proposes a data-driven probabilistic analysis approach for system risk assessment of the miniWECC system using actual data. The sparse Gaussian process (SGP) is advocated to quantify the impacts of uncertain inputs on voltage security. SGP does not need the probability distribution function of uncertain inputs, can handle correlations and is highly computationally efficient. Results on the miniWECC system using realistic data show that SGP outperforms existing approaches and is able to quantify the voltage violation risks.

17 WIND ENERGY↗
DOE OSTI2024

Software Verification of VARPOW

The VARPOW program is a post-processing utility program for DIF3D, specifically DIF3DVARIANT, and it was developed to provide interface files for the thermal analysis program DASSH. The basic methodology of VARPOW is to take the neutron and gamma flux (moments) calculated by DIF3D (or GAMSOR) and combine them with the heating (coefficient) cross sections to calculate the spatial power distributions using the DIF3DVARIANT spatial basis. VARPOW can use the output from GAMSOR (both steady state neutron and gamma flux calculations) or standard DIF3D/REBUS calculations (neutron flux only). The correct approach for defining the power distribution is to use GAMSOR as its purpose was to properly compute the gamma heating throughout the modeled domain. The power densities calculated by VARPOW are broken into fuel, cladding and coolant terms for which isotope-wise categorization is needed. VARPOW has built in options the user can select for the isotope categorization or VARPOW can import a file that details the isotope categorization. VARPOW can export the solution in the polynomial basis of DIF3D-VARIANT or the monomial basis of DIF3D-VARIANT. The purpose of this work is to verify the power distribution results calculated by VARPOW from both the GAMSOR and DIF3D input options and verify that the input and output options are consistent with the manual. Hand calculation and independent numerical calculation are used for this verification work.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Dynamic Model Agnostic Reliability Evaluation of Machine-Learning Models Integrated in Instrumentation & Control Systems

In recent years, the field of data-driven neural network-based machine learning (ML) algorithms has grown significantly and spurred research in its applicability to instrumentation and control systems. While they are promising in operational contexts, the trustworthiness of such algorithms is not adequately assessed. Failures of ML-integrated systems are poorly understood; the lack of comprehensive risk modeling can degrade the trustworthiness of these systems. In recent reports by the National Institute for Standards and Technology, trustworthiness in ML is a critical barrier to adoption and will play a vital role in intelligent systems' safe and accountable operation. Thus, in this work, we demonstrate a real-time model-agnostic method to evaluate the relative reliability of ML predictions by incorporating out-of-distribution detection on the training dataset. It is well documented that ML algorithms excel at interpolation (or near-interpolation) tasks but significantly degrade at extrapolation. This occurs when new samples are "far" from training samples. The method, referred to as the Laplacian distributed decay for reliability (LADDR), determines the difference between the operational and training datasets, which is used to calculate a prediction's relative reliability. LADDR is demonstrated on a feedforward neural network-based model used to predict safety significant factors during different loss-of-flow transients. LADDR is intended as a "data supervisor" and determines the appropriateness of well-trained ML models in the context of operational conditions. Ultimately, LADDR illustrates how training data can be used as evidence to support the trustworthiness of ML predictions when utilized for conventional interpolation tasks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Software Verification of VARPOW

The VARPOW program is a post-processing utility program for DIF3D, specifically DIF3DVARIANT, and it was developed to provide interface files for the thermal analysis program DASSH. The basic methodology of VARPOW is to take the neutron and gamma flux (moments) calculated by DIF3D (or GAMSOR) and combine them with the heating (coefficient) cross sections to calculate the spatial power distributions using the DIF3DVARIANT spatial basis. VARPOW can use the output from GAMSOR (both steady state neutron and gamma flux calculations) or standard DIF3D/REBUS calculations (neutron flux only). The correct approach for defining the power distribution is to use GAMSOR as its purpose was to properly compute the gamma heating throughout the modeled domain. The power desnities calculated by VARPOW are broken into fuel, cladding and coolant terms for which isotope-wise categorization is needed. VARPOW has built in options the user can select for the isotope categorization or VARPOW can import a file that details the isotope categorization. VARPOW can export the solution in the polynomial basis of DIF3D-VARIANT or the monomial basis of DIF3D-VARIANT. The purpose of this work is to verify the power distribution results calculated by VARPOW from both the GAMSOR and DIF3D input options and verify that the input and output options are consistent with the manual. Hand calculation and independent numerical calculation are used for this verification work.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Distributed out-of-memory NMF on CPU/GPU architectures

We propose an efficient distributed out-of-memory implementation of the non-negative matrix factorization (NMF) algorithm for heterogeneous high-performance-computing systems. The proposed implementation is based on prior work on NMFk, which can perform automatic model selection and extract latent variables and patterns from data. In this work, we extend NMFk by adding support for dense and sparse matrix operation on multi-node, multi-GPU systems. The resulting algorithm is optimized for out-of-memory problems where the memory required to factorize a given matrix is greater than the available GPU memory. Memory complexity is reduced by batching/tiling strategies, and sparse and dense matrix operations are significantly accelerated with GPU cores (or tensor cores when available). Input/output latency associated with batch copies between host and device is hidden using CUDA streams to overlap data transfers and compute asynchronously, and latency associated with collective communications (both intra-node and inter-node) is reduced using optimized NVIDIA Collective Communication Library (NCCL) based communicators. Benchmark results show significant improvement, from 32X to 76x speedup, with the new implementation using GPUs over the CPU-based NMFk. Good weak scaling was demonstrated on up to 4096 multi-GPU cluster nodes with approximately 25,000 GPUs when decomposing a dense 340 Terabyte-size matrix and an 11 Exabyte-size sparse matrix of density 10 -6 .

97 MATHEMATICS AND COMPUTING↗