SEARCH · Engineering Papers

Results for “robust clustering”

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 163 records · Page 9

All sources NASA NTRS DOE OSTI

NASA NTRS1992

An extended Lagrangian method

A unique formulation of describing fluid motion is presented. The method, referred to as 'extended Lagrangian method', is interesting from both theoretical and numerical points of view. The formulation offers accuracy in numerical solution by avoiding numerical diffusion resulting from mixing of fluxes in the Eulerian description. Meanwhile, it also avoids the inaccuracy incurred due to geometry and variable interpolations used by the previous Lagrangian methods. Unlike the Lagrangian method previously imposed which is valid only for supersonic flows, the present method is general and capable of treating subsonic flows as well as supersonic flows. The method proposed in this paper is robust and stable. It automatically adapts to flow features without resorting to clustering, thereby maintaining rather uniform grid spacing throughout and large time step. Moreover, the method is shown to resolve multi-dimensional discontinuities with a high level of accuracy, similar to that found in one-dimensional problems.

Liou, Meng-Sing↗

NASA NTRS1993

An extended Lagrangian method

A unique formulation of describing fluid motion is presented. The method, referred to as 'extended Lagrangian method', is interesting from both theoretical and numerical points of view. The formulation offers accuracy in numerical solution by avoiding numerical diffusion resulting from mixing of fluxes in the Eulerian description. Meanwhile, it also avoids the inaccuracy incurred due to geometry and variable interpolations used by the previous Lagrangian methods. The present method is general and capable of treating subsonic flows as well as supersonic flows. The method proposed in this paper is robust and stable. It automatically adapts to flow features without resorting to clustering, thereby maintaining rather uniform grid spacing throughout and large time step. Moreover, the method is shown to resolve multidimensional discontinuities with a high level of accuracy, similar to that found in 1D problems.

Liou, Meng-Sing↗

NASA NTRS1995

An Extended Lagrangian Method

A unique formulation of describing fluid motion is presented. The method, referred to as 'extended Lagrangian method,' is interesting from both theoretical and numerical points of view. The formulation offers accuracy in numerical solution by avoiding numerical diffusion resulting from mixing of fluxes in the Eulerian description. The present method and the Arbitrary Lagrangian-Eulerian (ALE) method have a similarity in spirit-eliminating the cross-streamline numerical diffusion. For this purpose, we suggest a simple grid constraint condition and utilize an accurate discretization procedure. This grid constraint is only applied to the transverse cell face parallel to the local stream velocity, and hence our method for the steady state problems naturally reduces to the streamline-curvature method, without explicitly solving the steady stream-coordinate equations formulated a priori. Unlike the Lagrangian method proposed by Loh and Hui which is valid only for steady supersonic flows, the present method is general and capable of treating subsonic flows and supersonic flows as well as unsteady flows, simply by invoking in the same code an appropriate grid constraint suggested in this paper. The approach is found to be robust and stable. It automatically adapts to flow features without resorting to clustering, thereby maintaining rather uniform grid spacing throughout and large time step. Moreover, the method is shown to resolve multi-dimensional discontinuities with a high level of accuracy, similar to that found in one-dimensional problems.

Liou, Meng-Sing↗

NASA NTRS2002

Gilgamesh: A Multithreaded Processor-In-Memory Architecture for Petaflops Computing

Processor-in-Memory (PIM) architectures avoid the von Neumann bottleneck in conventional machines by integrating high-density DRAM and CMOS logic on the same chip. Parallel systems based on this new technology are expected to provide higher scalability, adaptability, robustness, fault tolerance and lower power consumption than current MPPs or commodity clusters. In this paper we describe the design of Gilgamesh, a PIM-based massively parallel architecture, and elements of its execution model. Gilgamesh extends existing PIM capabilities by incorporating advanced mechanisms for virtualizing tasks and data and providing adaptive resource management for load balancing and latency tolerance. The Gilgamesh execution model is based on macroservers, a middleware layer which supports object-based runtime management of data and threads allowing explicit and dynamic control of locality and load balancing. The paper concludes with a discussion of related research activities and an outlook to future work.

management locality load balance↗

DOE OSTI2021

Cyber-Attack Detection for Photovoltaic Farms Based on Power-Electronics-Enabled Harmonic State Space Modeling

Here in this paper, a physics-data-based detection method is proposed to detect a variety of cyber-attacks in Photovoltaic (PV) farms using the power electronics-enabled harmonic state space (HSS) models, which, to our knowledge, is original. At the device level, HSS-based detection is developed to monitor harmonic vectors of individual PV converter with minimum sensor measurements, thus improving accuracy and robustness compared to Kalman Filter-based detection. At the system level that involves multiple PV converters, a clustering approach is developed to investigate attack propagation and accurately locate attack sources within a PV farm. The proposed approach is one of the first attempts to address PV security through interaction between the device and system, maximizing the accuracy and robustness at different levels. To verify the feasibility, a comprehensive attacks model is built, including single attack, coordinated attacks, and replay attacks. Besides, the impacts of irradiance changes are taken into consideration in the test scenarios. With the real-time data acquisition and hardware-in-the-loop testbed, comprehensive test results are provided to verify the feasibility of the proposed detection methodology.

42 ENGINEERING↗

DOE OSTI2026

Intelligent Experiments through Real-Time AI: Fast Data Processing and Autonomous Detector Control for High-Energy Nuclear Experiments

The aim of this project is to develop software and hardware for fast real-time data processing and autonomous detector control and calibration for the sPHENIX and the future EIC experiments. Below summarizes Georgia Tech team efforts in the past year: 1. We developed a real-time clustering algorithm and FPGA-based pipeline architecture for processing fired pixel data from ALPIDE sensors in sPHENIX experiments. Our Columnar Clustering Co-Design introduces a hardware-aware, stream-friendly approach that segments pixel data by column pairs using a Column Pair Clustering (CPC) strategy, followed by Cluster Stitching to merge adjacent subclusters. Implemented in Vitis HLS, the pipeline comprises five stages—read-in, subclustering, stitching, analysis, and write-out—connected by tagged HLS streams with custom end-of-event signaling for robust synchronization. We designed a pipelined dataflow model optimized for throughput, low latency, and minimal buffering, enabling scalable clustering across events of arbitrary size. Our system maintains spatial precision via center-of-mass and shape key extraction and efficiently handles edge cases such as fragmented or nested clusters. Compared against DBSCAN in both software and hardware, our approach demonstrates competitive performance under FPGA constraints. 2. We also conducted a comprehensive algorithm-to-hardware co-design of connected component analysis tailored for sPHENIX experiments, focusing on real-time, low-latency processing using FPGAs and High-Level Synthesis (HLS). Starting from a Python-based particle tracking pipeline, the team translated the core logic—graph traversal via DFS and Union-Find—into an HLS-compatible C++ model, replacing dynamic memory and recursion with static arrays and pipelined control flow. The final design includes a fully streamed and dataflow-compatible Union-Find kernel optimized across five iterations, incorporating loop pipelining, array partitioning, AXI/FIFO interface tuning, and function flattening. Experimental results show up to 14.8× speedup over the CPU baseline, reducing per-graph latency to 1.58 μs and demonstrating strong resource efficiency with only ~7k LUTs and zero BRAM usage. The design maintains functional correctness against the Python reference using a Python-based C-simulation framework and Mean Squared Error metrics. This work validates the potential of HLS-driven FPGA designs for edge-level HEP data acquisition, laying a scalable foundation for future integration with real-time detector pipelines and multi-graph processing systems.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2026

Many-body entanglement in solid-state emitters

The preparation and control of quantum states lie at the heart of quantum information science. Recent advances in solid-state quantum emitters (QEs) and nanophotonics have transformed the landscape of quantum photonic technologies, enabling scalable generation of quantum states of light and matter. A new frontier in solid-state quantum photonics is the engineering of many-body interactions between QEs and photons to achieve robust coherence and controllable many-body entanglement. These entangled states, including photonic graph and cluster states, superradiant emission and emergent quantum phases, are promising for quantum computation, sensing and simulation. However, intrinsic inhomogeneities and decoherence in solid-state platforms pose considerable challenges in realizing such complex entangled states. This Review provides an overview of fundamental many-body interactions and dynamics at the light–matter interfaces of solid-state QEs and discusses recent advances in mitigating decoherence and harnessing robust many-body coherence.

36 MATERIALS SCIENCE↗

DOE OSTI2025

The Effects of Compounded Model Size Reductions on Adversarial Robustness

Recent advances in Edge AI and Tiny Machine Learning (TinyML) have enabled the deployment of machine learning models on resource-constrained environments. However, deploying these models on edge devices, such as micro-controllers, requires significant model footprint reduction through a variety of techniques such as quantization, pruning, and clustering. While these optimization methods offer considerable advantages, they potentially introduce AI-related security vulnerabilities, particularly concerning model robustness with respect to adversarial AI attacks. Prior research has extensively examined the impact of quantization on adversarial robustness; however, the effects of alternative reduction techniques and their combinations remain understudied. This paper investigates the impact of model size reduction techniques on adversarial robustness, when applied individually and combined. We utilized Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) attacks to generate adversarial perturbations for both training and testing data, and then evaluated the models' accuracy under adversarial training conditions. Our findings revealed that reduction techniques generally diminished robustness; although, combining techniques was not found to make robustness any worse than when applied individually. Moreover, specific techniques can potentially enhance resistance to small size perturbations. This research provides insights into the trade-offs between model size reduction and security, establishing a foundation for future investigations into improving adversarial training techniques and methodologies for maintaining robustness while preserving memory footprint benefits.

Austria, Phillipe [ORNL] (ORCID:0000000236223973)↗

DOE OSTI2024

Analytic Gradients for Equation-of-Motion Coupled Cluster with Single, Double, and Perturbative Triple Excitations

Understanding the process of molecular photoexcitation is crucial in various fields, including drug development, materials science, photovoltaics, and more. The electronic vertical excitation energy is a critical property, for example in determining the singlet-triplet gap of chromophores. However, a full understanding of excited-state processes requires additional explorations of the excited-state potential energy surface and electronic properties, which is greatly aided by the availability of analytic energy gradients. Owing to its robust high accuracy over a wide range of chemical problems, equation-of-motion coupled-cluster with single and double excitations (EOM-CCSD) is a powerful method for predicting excited state properties, and the implementation of analytic gradients of many EOM-CCSD (excitation energies, ionization potentials, electron attachment energies, etc.) along with numerous successful applications high- lights the flexibility of the method. In specific cases where a higher level of accuracy is needed or in more complex electronic structures, the inclusion of triple excitations becomes essential, for example, in the EOM-CCSD* approach of Saeh and Stanton. In this work, we derive and implement for the first time the analytic gradients of EOMEE-CCSD*, which also provides a template for analytic gradients of related ex- cited state methods with perturbative triple excitations. Here, the capabilities of analytic EOMEE-CCSD* gradients are illustrated by several representative examples.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

An atomic cluster expansion potential for twisted multilayer graphene

Twisted multilayer graphene, characterized by its moiré patterns arising from inter-layer rotational misalignment, serves as a rich platform for exploring quantum phenomena. Machine learning interatomic potentials (MLIPs) are a promising approach to model such systems. Our work develops a method to generate training and test datasets for fitting MLIPs that capture all possible misalignments but remain small-scale to facilitate efficient data generation and parameter estimation. To achieve this, we generate configurations with periodic boundary conditions suitable for density functional theory calculations, and then introduce an internal twist and shift within those supercell structures. Using this technique, supplemented with an active learning workflow, we fit an Atomic Cluster Expansion potential for simulating twisted multilayer graphene and test it for accuracy and robustness on a range of simulation tasks.

2D materials↗

DOE OSTI2020

First-principles modeling of chemistry in mixed solvents: Where to go from here?

Mixed solvents (i.e., binary or higher order mixtures of ionic or nonionic liquids) play crucial roles in chemical syntheses, separations, and electrochemical devices because they can be tuned for specific reactions and applications. Apart from fully explicit solvation treatments that can be difficult to parameterize or computationally expensive, there is currently no well-established first-principles regimen for reliably modeling atomic-scale chemistry in mixed solvent environments. We offer our perspective on how this process could be achieved in the near future as mixed solvent systems become more explored using theoretical and computational chemistry. We first outline what makes mixed solvent systems far more complex compared to single-component solvents. An overview of current and promising techniques for modeling mixed solvent environments is provided. We focus on so-called hybrid solvation treatments such as the conductor-like screening model for real solvents and the reference interaction site model, which are far less computationally demanding than explicit simulations. We also propose that cluster-continuum approaches rooted in physically rigorous quasi-chemical theory provide a robust, yet practical, route for studying chemical processes in mixed solvents.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Dark Energy Survey Year 3 results: galaxy clustering and systematics treatment for lens galaxy samples

ABSTRACT In this work, we present the galaxy clustering measurements of the two DES lens galaxy samples: a magnitude-limited sample optimized for the measurement of cosmological parameters, maglim, and a sample of luminous red galaxies selected with the redmagic algorithm. maglim/redmagic sample contains over 10 million/2.5 million galaxies and is divided into six/five photometric redshift bins spanning the range z ∈ [0.20, 1.05]/z ∈ [0.15, 0.90]. Both samples cover 4143 $\deg ^2$ over which we perform our analysis blind, measuring the angular correlation function with an S/N ∼ 63 for both samples. In a companion paper, these measurements of galaxy clustering are combined with the correlation functions of cosmic shear and galaxy–galaxy lensing of each sample to place cosmological constraints with a 3 × 2pt analysis. We conduct a thorough study of the mitigation of systematic effects caused by the spatially varying survey properties and we correct the measurements to remove artificial clustering signals. We employ several decontamination methods with different configurations to ensure the robustness of our corrections and to determine the systematic uncertainty that needs to be considered for the final cosmology analyses. We validate our fiducial methodology using lognormal mocks, showing that our decontamination procedure induces biases no greater than 0.5σ in the (Ωm, b) plane, where b is the galaxy bias.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2020

Density weighted angular redshift fluctuations: a new cosmological observable

We propose the use of angular fluctuations in the galaxy redshift field as a new way to extract cosmological information in the Universe. This new probe $\delta z (\hat{\mathbf {n}})$ consists of the statistics of sky maps built by projecting redshifts under a Gaussian window of width σz centred upon a redshift zobs, and weighted by the galaxy density field. We compute the angular power spectrum of the $\delta z (\hat{\mathbf {n}})$ field in both numerical simulations and in linear perturbation theory. From these, we find that the $\delta z (\hat{\mathbf {n}})$ field (i) is sensitive to the underlying density and peculiar velocity fields; (ii) is highly correlated, at the $\gtrsim 60\, {{\ \rm per\ cent}}$ level, to the line-of-sight projected peculiar velocity field; (iii) for narrow windows (σz < 0.03), it is almost completely uncorrelated to the projected galaxy angular density field under the same redshift window; and (iv) it is largely unaffected by multiplicative and additive systematic errors on the observed number of galaxies that are redshift-independent over ~σz. We conclude that $\delta z (\hat{\mathbf {n}})$ is a simple and robust tomographic measure of the cosmic density and velocity fields, complementary to angular clustering, that will contribute to more complete exploitations of current and upcoming galaxy redshift surveys.

79 ASTRONOMY AND ASTROPHYSICS↗

NASA NTRS1994

Two-fluid simulations of galaxy formation

We investigate the formation of galaxies and larger structure with a simulation modeling two gravitationally coupled fluids representing dark matter and baryons. The baryon gas dynamics are calculated with a smoothed particle hydrodynamics (SPH) method, and the physics modeled includes thermal pressure, shock heating, and radiative cooling. We simulate a 16 Mpc periodic cube with 64(exp 3) particles in each fluid and 10% baryon mass fraction. We confirm, for the first time experimentally, disk formation as a natural consequence of hierarchical clustering in a large-scale cosmological environment. The majority of isolated galaxies exhibit centrifugally supported disks. A power-law relation between cold baryonic mass and maximum rotation velocity is found, M varies as nu(sub rot)(exp alpha) with alpha = 2.5 after correcting for differential numerical resolution. Both the spatial and velocity distributions of the simulated galaxies are biased with respect to the dark matter. A counts-in-cells analysis indicates that an unphysical degree of merging in the central cluster is likely responsible for the antibias signal in the correlation function. A robust, scale-dependent velocity bias is measured. The ratio of galaxy to dark matter pairwise velocity dispersions on a scale of 1 Mpc is 0.7. The amplitude is only mildly dependent on redshift or mass cutoff and scales with separation as r(exp 0.2). The degree to which these results depend on numerical parameters is discussed. Mass resolution plays a key role in controlling the resulting fraction of cold, dense baryons. The mass fraction associated with galaxies decreases by a factor of approximately greater than 3 when the mass per particle is increased by a factor 8. Photoionization and energy input from supernova will have to be included to determine more carefully the fraction of highly dissipated material and the characteristics of the stellar component of galaxies.

Evrard, August E.↗

NASA NTRS2015

Pendulum Motion in Main Parachute Clusters

The coupled dynamics of a cluster of parachutes to a payload are notoriously difficult to predict. Often the payload is designed to be insensitive to the range of attitude and rates that might occur, but spacecraft generally do not have the mass and volume budgeted for this robust of a design. The National Aeronautics and Space Administration (NASA) Orion Capsule Parachute Assembly System (CPAS) implements a cluster of three mains for landing. During testing of the Engineering Development Unit (EDU) design, it was discovered that with a cluster of two mains (a fault tolerance required for human rating) the capsule coupled to the parachute cluster could get into a limit cycle pendulum motion which would exceed the spacecraft landing capability. This pendulum phenomenon could not be predicted with the existing models and simulations. A three phased effort has been undertaken to understand the consequence of the pendulum motion observed, and explore potential design changes that would mitigate this phenomenon. This paper will review the early analysis that was performed of the pendulum motion observed during EDU testing, summarize the analysis ongoing to understand the root cause of the pendulum phenomenon, and discuss the modeling and testing that is being pursued to identify design changes that would mitigate the risk.

Ray, Eric S.↗

NASA NTRS1993

The dynamics of Abell 2634

We have amassed a large sample of velocity data for the cluster of galaxies Abell 2634 which contains the wide-angle tail (WAT) radio source 3C 465. Robust indicators of location and scale and their confidence intervals are used to determine if the cD galaxy, containing the WAT, has a significant peculiar motion. We find a cD peculiar radial velocity of 219 plus or minus 98 km s(exp -1). Further dynamical analyses, including substructure and normality tests, suggest that A 2634 is an unrelaxed cluster whose radio source structure may be bent by the turbulent gas of a recent cluster-subcluster merger.

Pinkney, J.↗

DOE OSTI2024

Dark energy survey year 3 results: miscentring calibration and X-ray-richness scaling relations in redMaPPer clusters

ABSTRACT We use Dark Energy Survey Year 3 (DES Y3) clusters with archival XMM–Newton and Chandra X-ray data to assess the centring performance of the redMaPPer cluster finder and to measure key richness observable scaling relations. We find that 10–20 per cent of redMaPPer clusters are miscentred, both when comparing to the X-ray peak position and to the visually identified central cluster galaxy. We find no significant difference in miscentring in bins of low versus high richness or redshift. The dominant reasons for miscentring include masked or missing data and the presence of other bright galaxies in the cluster. For half of the miscentred clusters, the correct central was one of the possible centrals identified by redMaPPer, while for ∼40 per cent of miscentred clusters, the correct central is not a redMaPPer member mostly due to masking. Additionally, we fit scaling relations of X-ray temperature and luminosity with richness. We find a TX–λ scatter of $0.21\pm 0.01$. While the scatter in TX–λ is consistent in redshift bins, we find modestly different slopes, with high-redshift clusters displaying a somewhat shallower relation. Splitting based on richness, we find a marginally larger scatter for our lowest richness bin, 20 < λ < 40. We note that the robustness of the scaling relations at lower richnesses is limited by the unknown selection function, but at λ > 75, we detect nearly all of the clusters falling within existing X-ray pointings. The X-ray properties of detected, serendipitous clusters are generally consistent with those of targeted clusters.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2022

Updated neutrino mass constraints from galaxy clustering and CMB lensing-galaxy cross-correlation measurements

In this work, we revisit cosmological constraints on the sum of the neutrino masses Σm v from a combination of full-shape BOSS galaxy clustering [P(k)] data and measurements of the cross-correlation between Planck Cosmic Microwave Background (CMB) lensing convergence and BOSS galaxy overdensity maps [$C_{l}^{kg}$], using a simple but theoretically motivated model for the scale-dependent galaxy bias in auto- and cross-correlation measurements. We improve upon earlier related work in several respects, particularly through a more accurate treatment of the correlation and covariance between P(k) and $C_{l}^{kg}$ measurements. When combining these measurements with Planck CMB data, we find a 95% confidence level upper limit of Σm v sum of neutrino masses < 0.14 eV, while slightly weaker limits are obtained when including small-scale ACTPol CMB data, in agreement with our expectations. We confirm earlier findings that (once combined with CMB data) the full-shape information content is comparable to the geometrical information content in the reconstructed BAO peaks given the precision of current galaxy clustering data, discuss the physical significance of our inferred bias and shot noise parameters, and perform a number of robustness tests on our underlying model. While the inclusion of $C_{l}^{kg}$ measurements does not currently appear to lead to substantial improvements in the resulting Σm v of neutrino mass constraints, we expect the converse to be true for near-future galaxy clustering measurements, whose shape information content will eventually supersede the geometrical one.

79 ASTRONOMY AND ASTROPHYSICS↗