Search NASA⌕ Search

SEARCH · Search NASA

Results for “benchmark problem”

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 1,801 records · Page 100

Open Rotor Tone Shielding Methods for System Noise Assessments Using Multiple Databases

Advanced aircraft designs such as the hybrid wing body, in conjunction with open rotor engines, may allow for significant improvements in the environmental impact of aviation. System noise assessments allow for the prediction of the aircraft noise of such designs while they are still in the conceptual phase. Due to significant requirements of computational methods, these predictions still rely on experimental data to account for the interaction of the open rotor tones with the hybrid wing body airframe. Recently, multiple aircraft system noise assessments have been conducted for hybrid wing body designs with open rotor engines. These assessments utilized measured benchmark data from a Propulsion Airframe Aeroacoustic interaction effects test. The measured data demonstrated airframe shielding of open rotor tonal and broadband noise with legacy F7/A7 open rotor blades. Two methods are proposed for improving the use of these data on general open rotor designs in a system noise assessment. The first, direct difference, is a simple octave band subtraction which does not account for tone distribution within the rotor acoustic signal. The second, tone matching, is a higher-fidelity process incorporating additional physical aspects of the problem, where isolated rotor tones are matched by their directivity to determine tone-by-tone shielding. A case study is conducted with the two methods to assess how well each reproduces the measured data and identify the merits of each. Both methods perform similarly for system level results and successfully approach the experimental data for the case study. The tone matching method provides additional tools for assessing the quality of the match to the data set. Additionally, a potential path to improve the tone matching method is provided.

Bahr, Christopher J.↗

Unsupervised Anomaly Detection in High-Dimensional Flight Data Using Convolutional Variational Auto-Encoder

The modern National Airspace System (NAS) is an extremely safe system. The industry has experienced a steady decrease in fatalities over the years. This can be contributed to both improved flight critical systems with redundant hardware and software protections as well as an increased focus on active monitoring and response to real time and historically identified vulnerabilities by implementing more resilient procedures and protocols. The main practice for identifying vulnerabilities in operations leverages domain expertise using knowledge about how the system should behave with the expected tolerances to known safety margins. This approach works well when the system has a well-defined operating condition. However, the operations in the NAS can be highly complex with various nuances that render it difficult to clearly pre-define all known safety vulnerabilities. With the advancement of data science and machine learning techniques, the potential to automatically identify emerging vulnerabilities in the observed operations has become more practical in recent years. The state-of-the-art anomaly detection approaches in aerospace data usually rely on supervised or semi-supervised learning. However, in many real-world problems such as flight safety creating labels for the data requires huge amount of efforts and is largely expensive. As a result, in this article, we develop a Convolutional Variational Auto-Encoder (CVAE), an unsupervised learning approach for anomaly detection in high-dimensional heterogeneous time-series data. We validate performance of CVAE compared to the state-of-the-art supervised learning approach (as an upper bound) as well as an supervised clustering based on K-Means (as a lower bound) on Yahoo!'s benchmark time series anomaly detection data. Finally, we showcase performance of CVAE on a case study of identifying anomalies in the first 60 seconds of commercial flights' take-offs using Flight Operational Quality Assurance (FOQA) data.

Milad Memarzadeh↗

Laboratory measurements of large-scale carbon sequestration flows in saline reservoirs

Brine saturated with CO{sub 2} is slightly denser than the original brine causing it to sink to the bottom of a saline reservoir where the CO{sub 2} is safely sequestered. However, the buoyancy of pure CO{sub 2} relative to brine drives it to the top of the reservoir where it collects underneath the cap rock as a separate phase of supercritical fluid. Without additional processes to mix the brine and CO{sub 2}, diffusion in this geometry is slow and would require an unacceptably long time to consume the pure CO{sub 2}. However, gravity and diffusion-driven convective instabilities have been hypothesized that generate enhanced CO{sub 2}-brine mixing promoting dissolution of CO{sub 2} into the brine on time scale of a hundred years. These flows involve a class of hydrodynamic problems that are notoriously difficult to simulate; the simultaneous flow of mUltiple fluids (CO{sub 2} and brine) in porous media (rock or sediment). The hope for direct experimental confirmation of simulations is dim due to the difficulty of obtaining high resolution data from the subsurface and the high pressures ({approx}100 bar), long length scales ({approx}100 meters), and long time scales ({approx}100 years) that are characteristic of these flows. We have performed imaging and mass transfer measurements in similitude-scaled laboratory experiments that provide benchmarks to test reservoir simulation codes and enhance their predictive power.

58 GEOSCIENCES↗

Dense LU Factorization on Multicore Supercomputer Nodes

Dense LU factorization is a prominent benchmark used to rank the performance of supercomputers. Many implementations, including the reference code HPL, use block-cyclic distributions of matrix blocks onto a two-dimensional process grid. The process grid dimensions drive a trade-off between communication and computation and are architecture- and implementation-sensitive. We show how the critical panel factorization steps can be made less communication-bound by overlapping asynchronous collectives for pivot identification and exchange with the computation of rank-k updates. By shifting this trade-off, a modified block-cyclic distribution can beneficially exploit more available parallelism on the critical path, and reduce panel factorization's memory hierarchy contention on now-ubiquitous multi-core architectures. The missed parallelism in traditional block-cyclic distributions arises because active panel factorization, triangular solves, and subsequent broadcasts are spread over single process columns or rows (respectively) of the process grid. Increasing one dimension of the process grid decreases the number of distinct processes in the other dimension. To increase parallelism in both dimensions, periodic 'rotation' is applied to the process grid to recover the row-parallelism lost by a tall process grid. During active panel factorization, rank-1 updates stream through memory with minimal reuse. In a column-major process grid, the performance of this access pattern degrades as too many streaming processors contend for access to memory. A block-cyclic mapping in the more popular row-major order does not encounter this problem, but consequently sacrifices node and network locality in the critical pivoting steps. We introduce 'striding' to vary between the two extremes of row- and column-major process grids. As a test-bed for further mapping experiments, we describe a dense LU implementation that allows a block distribution to be defined as a general function of block to processor. Other mappings can be tested with only small, local changes to the code.

99 GENERAL AND MISCELLANEOUS↗

The value of protein structure classification information-Surveying the scientific literature

The Structural Classification of Proteins (SCOP) and Class, Architecture, Topology, Homology (CATH) databases have been valuable resources for protein structure classification for over 20 years. Development of SCOP (version 1) concluded in June 2009 with SCOP 1.75. The SCOPe (SCOP-extended) database offers continued development of the classic SCOP hierarchy, adding over 33,000 structures. We have attempted to assess the impact of these two decade old resources and guide future development. To this end, we surveyed recent articles to learn how structure classification data are used. Of 571 articles published in 2012-2013 that cite SCOP, 439 actually use data from the resource. We found that the type of use was fairly evenly distributed among four top categories: A) study protein structure or evolution (27% of articles), B) train and/or benchmark algorithms (28% of articles), C) augment non-SCOP datasets with SCOP classification (21% of articles), and D) examine the classification of one protein/a small set of proteins (22% of articles). Most articles described computational research, although 11% described purely experimental research, and a further 9% included both. We examined how CATH and SCOP were used in 158 articles that cited both databases: while some studies used only one dataset, the majority used data from both resources. Protein structure classification remains highly relevant for a diverse range of problems and settings.

59 BASIC BIOLOGICAL SCIENCES↗

A novel algorithm of immersed moving boundary scheme for fluid-particle interactions in DEMLBM

This paper presents an efficient and accurate Immersed Moving Boundary (IMB) algorithm for solving fluid-particle interactions in the framework of the Lattice Boltzmann Method (LBM). Although the IMB scheme has been widely employed in many fluid-particle coupling problems in a wide range of applications, the algorithm of its implementation, especially in identifying both fluid and solid boundary nodes for particles, is seldom reported. Besides, the computational cost of handling fluid-particle coupling is very expensive. To provide a bridge between theory and application and improve the computing efficiency of IMB, a novel boundary tracing procedure and an efficient method for computing the solid nodal ratio using Gaussian quadrature are proposed in this paper. Both accuracy and efficiency of the proposed algorithm are examined by two benchmark tests. It is also found that the IMB scheme are more efficient and stable compared to another widely used the Immersed Boundary Method (IBM) in LBM.

42 ENGINEERING↗

DISCOVERY OF A HIGHLY UNEQUAL-MASS BINARY T DWARF WITH KECK LASER GUIDE STAR ADAPTIVE OPTICS: A COEVALITY TEST OF SUBSTELLAR THEORETICAL MODELS AND EFFECTIVE TEMPERATURES

Highly unequal-mass ratio binaries are rare among field brown dwarfs, with the mass ratio distribution of the known census described by q {sup (4.9{+-}0.7)}. However, such systems enable a unique test of the joint accuracy of evolutionary and atmospheric models, under the constraint of coevality for the individual components (the 'isochrone test'). We carry out this test using two of the most extreme field substellar binaries currently known, the T1 + T6 {epsilon} Ind Bab binary and a newly discovered 0.''14 T2.0 + T7.5 binary, 2MASS J12095613-1004008AB, identified with Keck laser guide star adaptive optics. The latter is the most extreme tight binary resolved to date (q {approx} 0.5). Based on the locations of the binary components on the Hertzsprung-Russell (H-R) diagram, current models successfully indicate that these two systems are coeval, with internal age differences of log(age) = -0.8 {+-} 1.3(-1.0{sup +1.2}{sub -1.3}) dex and 0.5{sup +0.4}{sub -0.3}(0.3{sup +0.3}{sub -0.4}) dex for 2MASS J1209-1004AB and {epsilon} Ind Bab, respectively, as inferred from the Lyon (Tucson) models. However, the total mass of {epsilon} Ind Bab derived from the H-R diagram ({approx} 80 M{sub Jup} using the Lyon models) is strongly discrepant with the reported dynamical mass. This problem, which is independent of the assumed age of the {epsilon} Ind Bab system, can be explained by a {approx} 50-100 K systematic error in the model atmosphere fitting, indicating slightly warmer temperatures for both components; bringing the mass determinations from the H-R diagram and the visual orbit into consistency leads to an inferred age of {approx} 6 Gyr for {epsilon} Ind Bab, older than previously assumed. Overall, the two T dwarf binaries studied here, along with recent results from T dwarfs in age and mass benchmark systems, yield evidence for small ({approx}100 K) errors in the evolutionary models and/or model atmospheres, but not significantly larger. Future parallax, resolved spectroscopy, and dynamical mass measurements for 2MASS J1209-1004AB will enable a more stringent application of the isochrone test. Finally, the binary nature of this object reduces its utility as the primary T3 near-IR spectral typing standard; we suggest SDSS J1206+2813 as a replacement.

79 ASTRONOMY AND ASTROPHYSICS↗

Applying Graph Partitioning Methods in Measurement-Based Dynamic Load Balancing

Load imbalance can lead to wasted CPU hours especially when running on a large number of processors. Achieving the best parallel efficiency for a program requires optimal load balancing which is a NP-hard problem. Charm++, a migratable objects based programming model, provides a measurement-based dynamic load balancing frame work. This paper explores the use of graph partitioning algorithms, traditionally used for partitioning physical domains/meshes, for measurement based dynamic load balancing of parallel applications. In particular, we present repartitioning methods developed in a graph partitioning toolbox called Scotch that consider the previous mapping to minimize migration costs. We also discuss a new imbalance reduction algorithm for graphs with irregular load distributions. We compare several load balancing algorithms using a micro-benchmark and the NAS BT multi-zone bench mark. New algorithms developed in Scotch lead to better performance compared to other existing partitioners, both in terms of the application execution time and fewer number of objects migrated.

97 MATHEMATICS AND COMPUTING↗

Scalable Predictive Control and Optimization for Grid Integration of Large-Scale Distributed Energy Resources: Preprint

Integration of a large number of distributed energy resources (DERs) into the power grid needs a scalable power balancing method. We formulate the power balancing problem as a look-ahead optimization problem to be solved sequentially by a power distribution system aggregator based on a model predictive control (MPC) framework. Solving large-scale look-ahead control problem requires proper configuration of the control steps. In this paper, to solve large-scale control problems, we propose a variable time granularity where control time steps nearby the current control step have finer resolutions. The aggregator objective includes maximization of power production revenue and minimization of power purchasing expense, renewable power curtailment, and mileage costs for energy storage and electric vehicle (EV) charging stations while satisfying system capacity and operational constraints. The control problem is formulated as a mixed-integer linear program (MILP) and solved using the XpressMP solver. We perform simulations considering a copper plate representation of a large distribution network consisting of 2507 devices (controllable DERs) including curtailable photovoltaics (PVs), energy storage batteries, EV charging stations, and buildings with heating, ventilation, and air conditioning units (HVACs). We show the effectiveness of the proposed approach in managing DERs interactively for maximum energy trading profit and local supply-demand power balancing. Finally, we demonstrate that the proposed method outperformed other benchmark controllers regarding computation time without compromising operational performance.

DER↗

Scalable Predictive Control and Optimization for Grid Integration of Large-Scale Distributed Energy Resources

Integrating a large number of distributed energy resources (DERs) into the power grid needs a scalable power balancing method. We formulate the power balancing problem as a look-ahead optimization problem to be solved sequentially by a power distribution system aggregator based on a model predictive control (MPC) framework. Solving large-scale look-ahead control problems requires proper configuration of the control steps. In this paper, to solve large-scale control problems, we propose a variable time granularity where control time steps nearby the current control step have finer resolutions. The aggregator objective includes maximization of power production revenue and minimization of power purchasing expense, renewable power curtailment, and mileage costs for energy storage and electric vehicle (EV) charging stations while satisfying system capacity and operational constraints. The control problem is formulated as a mixed-integer linear program (MILP) and solved using the XpressMP solver. We perform simulations considering a copper plate representation of a large distribution network consisting of 2507 devices (controllable DERs), including curtailable photovoltaics (PVs), energy storage batteries, EV charging stations, and buildings with heating, ventilation, and air conditioning units (HVACs). We show the effectiveness of the proposed approach in managing DERs interactively for maximum energy trading profit and local supply-demand power balancing. Finally, we demonstrate that the proposed method outperforms other benchmark controllers regarding computation time without compromising operational performance.

DER↗

Direct Estimation of Parameters in ODE Models Using WENDy: Weak-Form Estimation of Nonlinear Dynamics

Abstract We introduce the Weak-form Estimation of Nonlinear Dynamics (WENDy) method for estimating model parameters for non-linear systems of ODEs. Without relying on any numerical differential equation solvers, WENDy computes accurate estimates and is robust to large (biologically relevant) levels of measurement noise. For low dimensional systems with modest amounts of data, WENDy is competitive with conventional forward solver-based nonlinear least squares methods in terms of speed and accuracy. For both higher dimensional systems and stiff systems, WENDy is typically both faster (often by orders of magnitude) and more accurate than forward solver-based approaches. The core mathematical idea involves an efficient conversion of the strong form representation of a model to its weak form, and then solving a regression problem to perform parameter inference. The core statistical idea rests on the Errors-In-Variables framework, which necessitates the use of the iteratively reweighted least squares algorithm. Further improvements are obtained by using orthonormal test functions, created from a set of $$C^{\infty }$$ C ∞ bump functions of varying support sizes.We demonstrate the high robustness and computational efficiency by applying WENDy to estimate parameters in some common models from population biology, neuroscience, and biochemistry, including logistic growth, Lotka-Volterra, FitzHugh-Nagumo, Hindmarsh-Rose, and a Protein Transduction Benchmark model. Software and code for reproducing the examples is available at https://github.com/MathBioCU/WENDy .

97 MATHEMATICS AND COMPUTING↗

A new control score concept for building performance assessment

In buildings, performance assessment often focuses on energy use with metrics such as energy use intensity (EUI) used to benchmark performance. However, energy performance of a building is fundamentally determined by the control system that engages the energy-using systems. There are two aspects of control that are of particular importance: (1) the ability to regulate process variables to their setpoints; and (2) whether the setpoints are at the right levels and/or following desired profiles. Most buildings do not reach their energy efficiency potential due to deficiencies in control performance and operators do not have access to metrics that can illuminate these deficiencies. Here this paper addresses this problem by providing novel techniques that combine these two aspects of control performance into a single standardized score on the scale of 0-10. The concept of a standardized control scores enables all systems in a building to be compared on the same scale and also for scores to be rolled up to different levels in the building and system hierarchy for system-wide analysis. The paper presents the theory for the method, describes a prototype tool for displaying scores, and presents results from application to a large building in Minneapolis.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

The GLAS Algorithm Theoretical Basis Document for Precision Orbit Determination (POD)

The Geoscience Laser Altimeter System (GLAS) was the sole instrument for NASA's Ice, Cloud and land Elevation Satellite (ICESat) laser altimetry mission. The primary purpose of the ICESat mission was to make ice sheet elevation measurements of the polar regions. Additional goals were to measure the global distribution of clouds and aerosols and to map sea ice, land topography and vegetation. ICESat was the benchmark Earth Observing System (EOS) mission to be used to determine the mass balance of the ice sheets, as well as for providing cloud property information, especially for stratospheric clouds common over polar areas. The GLAS instrument operated from 2003 to 2009 and provided multi-year elevation data needed to determine changes in sea ice freeboard, land topography and vegetation around the globe, in addition to elevation changes of the Greenland and Antarctic ice sheets. This document describes the Precision Orbit Determination (POD) algorithm for the ICESat mission. The problem of determining an accurate ephemeris for an orbiting satellite involves estimating the position and velocity of the satellite from a sequence of observations. The ICESatGLAS elevation measurements must be very accurately geolocated, combining precise orbit information with precision pointing information. The ICESat mission POD requirement states that the position of the instrument should be determined with an accuracy of 5 and 20 cm (1-s) in radial and horizontal components, respectively, to meet the science requirements for determining elevation change.

Rim, Hyung Jin↗

Toward a Holistic Performance Evaluation of Large Language Models Across Diverse AI Accelerators

Artificial intelligence (AI) methods have become critical in scientific applications to help accelerate scientific discovery. Large language models (LLMs) are being considered a promising approach to address some challenging problems because of their superior generalization capabilities across domains. The effectiveness of the models and the accuracy of the applications are contingent upon their efficient execution on the underlying hardware infrastructure. Specialized Al accelerator hardware systems have recently become available for accelerating Al applications. However, the comparative performance of these AI accelerators on large language models has not been previously studied. In this paper, we systematically study LLMs on multiple AI accelerators and GPUs and evaluate their performance characteristics for these models. We evaluate these systems with (i) a micro-benchmark using a core transformer block, (ii) a GPT-2 model, and (iii) an 1,I,M-driven science use case, GenSLM. We present our findings and analyses of the models' performance to better understand the intrinsic capabilities of AI accelerators. Furthermore, our analysis takes into account key factors such as sequence lengths, scaling behavior, and sensitivity to gradient accumulation steps.

Emani, Murali↗

Scalable Predictive Control and Optimization for Grid Integration of Large-Scale Distributed Energy Resources

Integrating a large number of distributed energy resources (DERs) into the power grid needs a scalable power balancing method. We formulate the power balancing problem as a look-ahead optimization problem to be solved sequentially by a power distribution system aggregator based on a model predictive control (MPC) framework. Solving large-scale look-ahead control problems requires proper configuration of the control steps. In this paper, to solve large-scale control problems, we propose a variable time granularity where control time steps nearby the current control step have finer resolutions. The aggregator objective includes maximization of power production revenue and minimization of power purchasing expense, renewable power curtailment, and mileage costs for energy storage and electric vehicle (EV) charging stations while satisfying system capacity and operational constraints. The control problem is formulated as a mixed-integer linear program (MILP) and solved using the XpressMP solver. We perform simulations considering a copper plate representation of a large distribution network consisting of 2507 devices (control-lable DERs), including curtailable photovoltaics (PVs), energy storage batteries, EV charging stations, and buildings with heating, ventilation, and air conditioning units (HVACs). We show the effectiveness of the proposed approach in managing DERs interactively for maximum energy trading profit and local supply-demand power balancing. Finally, we demonstrate that the proposed method outperforms other benchmark controllers regarding computation time without compromising operational performance.

DER↗

Multiclass Classification Using Bayesian Multivariate Adaptive Regression Splines

We present a new Bayesian model for the problem of multiclass classification. In this model, the probabilities of class membership of a given observation are determined by the mean of a latent Gaussian distribution. The mean functions of this latent distribution consist of combinations of highly flexible basis functions of the inputs: multivariate adaptive regression splines (MARS), first developed for multiple regression. We use reversible jump Markov chain Monte Carlo to make inference on the classification model, including the number of basis functions. We compare the probabilistic classification performance of our proposed approach to existing methods on simulated and benchmark data, and compare uncertainty estimates on simulated data. Our proposed method compares favorably with existing Bayesian and frequentist multiclass classification methods in out-of-sample probabilistic classification, and uncertainty estimation of these probabilistic classifications. We examine the fit of the proposed method to a data set of hurricane storm surge levels near Delaware Bay, US, and conclude that sea level rise is a key contributor to damage delivered by storm surge.

97 MATHEMATICS AND COMPUTING↗

Least Squares Reverse Time Migration (LSRTM) for Damage Imaging in Composite Laminates

A method for adapting least squares reverse time migration (LSRTM) for ultrasonic guided wave imaging of composite laminates is proposed in this paper. As composites become more widely used in fields such as the aerospace industry, the need for high-resolution imaging in structural health monitoring (SHM) and nondestructive evaluation (NDE) is also growing. For instance, delamination is a common problem in composite laminates, which has led to a certain degree of apprehension in the use of composite materials for load-bearing structures. Although the solver-based imaging techniques using conventional reverse time migration (RTM) methods illuminate damage with a wide range of damage-scattering effects, the resulting images do not fully define the damage regions due to the limited data acquisition aperture, sensor density, frequencies/wavelengths, and incompleteness of adjoint reconstruction. Previously, the LSRTM theory was derived by incorporating a least squares migration (LSM) algorithm into RTM, and benchmarked its high-resolution damage imaging performance for isotropic plates. To improve damage imaging in composite laminates, a process to create an ultrasonic guided wave-based LSRTM method for anisotropic materials is proposed in this paper. The derivation of the forward modeling operator and the adjoint operator is presented. Numerical case studies were conducted to show the improvement of LSRTM over RTM in mapping damage in composite plates. Multiple damage sites or damage with a complex shape were created in the numerical studies based on 1) Born approximation-based modeling or 2) reduction of the plate stiffness-based modeling. LSRTM begins by using a forward modeling process. The forward modeling accuracy for the flexural wave solver used in this study for LSRTM was also compared with the results from a three-dimensional rotated staggered grid (RSG) elastodynamic finite difference solver, previously open-sourced by NASA Langley Research Center. These benchmarking studies have shown that the LSRTM method is more effective in reducing artifacts, improving resolution, and enhancing damage model reconstruction compared to RTM in composite laminates.

Jiaze He↗

Exact few-body results for strongly correlated quantum gases in two dimensions

The study of strongly correlated quantum gases in two dimensions has important ramifications for understanding many intriguing phenomena in solid materials, such as high-T{sub c} superconductivity and the fractional quantum-Hall effect. However, theoretical methods are plagued by the existence of significant quantum fluctuations. Here, we present two- and three-body exact solutions for both fermions and bosons trapped in a two-dimensional harmonic potential with an arbitrary s-wave scattering length. These few-particle solutions link in a natural way to the high-temperature properties of many-particle systems via a quantum virial expansion. As a concrete example, using the energy spectrum of few fermions, we calculate the second and third virial coefficients of a strongly interacting Fermi gas in two dimensions, and consequently investigate its high-temperature thermodynamics. Our thermodynamic results may be useful for ongoing experiments on two-dimensional Fermi gases. These exact results also provide an unbiased benchmark for quantum Monte Carlo simulations of two-dimensional Fermi gases at high temperatures.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Refine your search to explore more results.