Search NASA⌕ Search

SEARCH · Search NASA

Results for “Network parameter error”

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 199 records · Page 11

Photometric redshifts and clustering of emission line galaxies selected jointly by DES and eBOSS

We present the results of the first test plates of the extended Baryon Oscillation Spectroscopic Survey. This paper focuses on the emission line galaxies (ELG) population targetted from the Dark Energy Survey (DES) photometry. We analyse the success rate, efficiency, redshift distribution, and clustering properties of the targets. From the 9000 spectroscopic redshifts targetted, 4600 have been selected from the DES photometry. The total success rate for redshifts between 0.6 and 1.2 is 71\% and 68\% respectively for a bright and faint, on average more distant, samples including redshifts measured from a single strong emission line. We find a mean redshift of 0.8 and 0.87, with 15 and 13\% of unknown redshifts respectively for the bright and faint samples. In the redshift range 0.6<z<1.2, for the most secure spectroscopic redshifts, the mean redshift for the bright and faint sample is 0.85 and 0.9 respectively. Star contamination is lower than 2\%. We measure a galaxy bias averaged on scales of 1 and 10~Mpc/h of 1.72 \pm 0.1 for the bright sample and of 1.78 \pm 0.12 for the faint sample. The error on the galaxy bias have been obtained propagating the errors in the correlation function to the fitted parameters. This redshift evolution for the galaxy bias is in agreement with theoretical expectations for a galaxy population with MB-5\log h < -21.0. We note that biasing is derived from the galaxy clustering relative to a model for the mass fluctuations. We investigate the quality of the DES photometric redshifts and find that the outlier fraction can be reduced using a comparison between template fitting and neural network, or using a random forest algorithm.

79 ASTRONOMY AND ASTROPHYSICS↗

Short-Term Distribution System State Forecast Based on Optimal Synchrophasor Sensor Placement and Extreme Learning Machine

This paper proposes an approach for distribution system state forecasting, which aims to provide an accurate and high speed state forecasting with an optimal synchrophasor sensor placement (OSSP) based state estimator and an extreme learning machine (ELM) based forecaster. Specifically, considering the sensor installation cost and measurement error, an OSSP algorithm is proposed to reduce the number of synchrophasor sensor and keep the whole distribution system numerically and topologically observable. Then, the weighted least square (WLS) based system state estimator is used to produce the training data for the proposed forecaster. Traditionally, the artificial neural network (ANN) and support vector regression (SVR) are widely used in forecasting due to their nonlinear modeling capabilities. However, the ANN contains heavy computation load and the best parameters for SVR are difficult to obtain. In this paper, the ELM, which overcomes these drawbacks, is used to forecast the future system states with the historical system states. The proposed approach is effective and accurate based on the testing results.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Graybox and adaptative dynamic neural network identification models to infer the steady state efficiency of solar thermal collectors starting from the transient condition

This paper deals with the development of methods for non steady state test of solar thermal collectors. Our goal is to infer performances in steady-state conditions in terms of the efficiency curve when measures in transient conditions are the only ones available. We take into consideration the method of identification of a system in dynamic conditions by applying a Graybox Identification Model and a Dynamic Adaptative Linear Neural Network (ALNN) model. The study targets the solar collector with evacuated pipes, such as Dewar pipes. The mathematical description that supervises the functioning of the solar collector in transient conditions is developed using the equation of the energy balance, with the aim of determining the order and architecture of the two models. The input and output vectors of the two models are constructed, considering the measures of 4 days of solar radiation, flow mass, environment and heat-transfer fluid temperature in the inlet and outlet from the thermal solar collector. The efficiency curves derived from the two models are detected in correspondence to the test and validation points. The two synthetic simulated efficiency curves are compared with the actual efficiency curve certified by the Swiss Institute Solartechnik Puffung Forschung which tested the solar collector performance in steady-state conditions according to the UNI-EN 12975 standard. An acquisition set of measurements of only 4 days in the transient condition was enough to trace through a Graybox State Space Model the efficiency curve of the tested solar thermal collector, with a relative error of synthetic values with respect to efficiency certified by SPF, lower than 0.5%, while with the ALNN model the error is lower than 2.2% with respect to certified one. (author)

14 SOLAR ENERGY↗

Uncertainty propagation in feed-forward neural network models

We develop new uncertainty propagation methods for feed-forward neural network architectures with leaky ReLU activation functions subject to random perturbations in the input vectors. In particular, we derive analytical expressions for the probability density function (PDF) of the neural network output and its statistical moments as a function of the input uncertainty and the parameters of the network, i.e., weights and biases. A key finding is that an appropriate linearization of the leaky ReLU activation function yields accurate statistical results even for large perturbations in the input vectors. This can be attributed to the way information propagates through the network. We also propose new analytically tractable Gaussian copula surrogate models to approximate the full joint PDF of the neural network output. To validate our theoretical results, we conduct Monte Carlo simulations and a thorough error analysis on a multi-layer neural network representing a nonlinear integro-differential operator between two polynomial function spaces. Our findings demonstrate excellent agreement between the theoretical predictions and Monte Carlo simulations.

MLP networks↗

From Escherichia coli mutant 13 C labeling data to a core kinetic model: A kinetic model parameterization pipeline

Kinetic models of metabolic networks offer the promise of quantitative phenotype prediction. The mechanistic characterization of enzyme catalyzed reactions allows for tracing the effect of perturbations in metabolite concentrations and reaction fluxes in response to genetic and environmental perturbation that are beyond the scope of stoichiometric models. In this study, we develop a two-step computational pipeline for the rapid parameterization of kinetic models of metabolic networks using a curated metabolic model and available 13 C-labeling distributions under multiple genetic and environmental perturbations. The first step involves the elucidation of all intracellular fluxes in a core model of E. coli containing 74 reactions and 61 metabolites using 13 C-Metabolic Flux Analysis ( 13 C-MFA). Here, fluxes corresponding to the mid-exponential growth phase are elucidated for seven single gene deletion mutants from upper glycolysis, pentose phosphate pathway and the Entner-Doudoroff pathway. The computed flux ranges are then used to parameterize the same (i.e., k-ecoli74) core kinetic model for E. coli with 55 substrate-level regulations using the newly developed K-FIT parameterization algorithm. The K-FIT algorithm employs a combination of equation decomposition and iterative solution techniques to evaluate steady-state fluxes in response to genetic perturbations. k-ecoli74 predicted 86% of flux values for strains used during fitting within a single standard deviation of 13 C-MFA estimated values. By performing both tasks using the same network, errors associated with lack of congruity between the two networks are avoided, allowing for seamless integration of data with model building. Product yield predictions and comparison with previously developed kinetic models indicate shifts in flux ranges and the presence or absence of mutant strains delivering flux towards pathways of interest from training data significantly impact predictive capabilities. Using this workflow, the impact of completeness of fluxomic datasets and the importance of specific genetic perturbations on uncertainties in kinetic parameter estimation are evaluated.

59 BASIC BIOLOGICAL SCIENCES↗

Application of machine learning in the determination of impact parameter in the 132 Sn+ 124 Sn system

Here, 132 Sn + 124 Sn collisions at a beam energy of 270 MeV/nucleon were performed at the Radioactive Isotope Beam Factory (RIBF) in RIKEN to investigate the nuclear equation of state. Reconstructing the impact parameter is one of the important tasks in the experiment as it relates to many observable. In this work, we employ three commonly used algorithms in machine learning, the artificial neural network (ANN), the convolutional neural network (CNN), and the light gradient boosting machine (LightGBM), to determine the impact parameter by analyzing either the charged particle spectra or several features simulated with events from the ultrarelativistic quantum molecular dynamics (UrQMD) model. To closely imitate experimental data and investigate the generalizability of the trained machine learning algorithms, incompressibility of nuclear equation of state and the in-medium nucleon-nucleon cross sections are varied in the UrQMD model to generate the training data. The mean absolute error Δb between the true and the predicted impact parameter is smaller than 0.45 fm if training and testing sets are sampled from the UrQMD model with the same parameter set. However, if training and testing sets are sampled with different parameter sets, Δb would increase to 0.8 fm. The generalizability of the trained machine learning algorithms suggests that these machine learning algorithms can be used reliably to reconstruct the impact parameter in experiment.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulation and fitting of complex reaction network TPR: The key is the objective function

In this research, a method has been developed for finding improved fits during simulation and fitting of data from complex reaction network temperature programmed reactions (CRN-TPR). It was found that simulation and fitting of CRN-TPR presents additional challenges relative to simulation and fitting of simpler TPR systems. The method used here can enable checking the plausibility of proposed chemical mechanisms and kinetic models. The most important finding was that when choosing an objective function, use of an objective function that is based on integrated production provides more utility in finding improved fits when compared to an objective function based on the rate of production. The response surface produced by using the integrated production is monotonic, suppresses effects from experimental noise, requires fewer points to capture the response behavior, and can be simulated numerically with smaller errors. For CRN-TPR, there is increased importance (relative to simple reaction network TPR) in resolving of peaks prior to fitting, as well as from weighting of experimental data points. Using an implicit ordinary differential equation solver was found to be inadequate for simulating CRN-TPR. Lastly, the method employed here was capable of attaining improved fits in simulation and fitting of CRN-TPR when starting with a postulated mechanism and physically realistic initial guesses for the kinetic parameters.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

GAINN: The Galaxy Assembly and Interaction Neural Networks for High-redshift JWST Observations

We present the Galaxy Assembly and Interaction Neural Networks (Gainn), a series of artificial neural networks for predicting the redshift, stellar mass, halo mass, and mass-weighted age of simulated galaxies based on James Webb Space Telescope (JWST) photometry. Our goal is to determine the best neural network for predicting these variables at 11 < z < 15. The parameters of the optimal neural network can then be used to estimate these variables for real, observed galaxies. The inputs of the neural networks are JWST filter magnitudes of a subset of five broadband filters (F150W, F200W, F277W, F356W, and F444W) and two medium-band filters (F162M and F182M). We compare the performance of the neural networks using different combinations of these filters, as well as different activation functions and numbers of layers. The best neural network predicted redshift with a normalized rms error of $0.010^{+0.003}_{-0.001}$, stellar mass with rms = $0.089^{+0.044}_{-0.022}$, halo mass with a mean-squared error of $0.022^{+0.014}_{-0.008}$, and mass-weighted age with rms = $12.466^{+5.065}_{-2.408}$. We also test the performance of Gainn on real data from MACS0647JD, an object observed by JWST. Predictions from Gainn for the first projection of the object (JD1) have normalized bias $\langle$Δz$\rangle$ < 0.00228, which is significantly smaller than found with template-fitting methods. We find that the optimal filter combination is F277W, F356W, F162M, and F200W when considering both theoretical accuracy and observational resources from JWST.

97 MATHEMATICS AND COMPUTING↗

Smart Pixels: towards on-sensor inference of charged particle track parameters and uncertainties

The combinatorics of track seeding has long been a computational bottleneck for triggering and offline computing in High Energy Physics (HEP), and remains so for the HL-LHC. Next-generation pixel sensors will be sufficiently fine-grained to determine angular information of the charged particle passing through from pixel-cluster properties. This detector technology immediately improves the situation for offline tracking, but any major improvements in physics reach are unrealized since they are dominated by lowest-level hardware trigger acceptance. We will demonstrate track angle and hit position prediction, including errors, using a mixture density network within a single layer of silicon as well as the progress towards and status of implementing the neural network in hardware on both FPGAs and ASICs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Smart Pixels: towards on-sensor inference of charged particle track parameters and uncertainties

The combinatorics of track seeding has long been a computational bottleneck for triggering and offline computing in High Energy Physics (HEP), and remains so for the HL-LHC. Next-generation pixel sensors will be sufficiently fine-grained to determine angular information of the charged particle passing through from pixel-cluster properties. This detector technology immediately improves the situation for offline tracking, but any major improvements in physics reach are unrealized since they are dominated by lowest-level hardware trigger acceptance. We will demonstrate track angle and hit position prediction, including errors, using a mixture density network within a single layer of silicon as well as the progress towards and status of implementing the neural network in hardware on both FPGAs and ASICs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Smartpixels: Towards on-sensor inference of charged particle track parameters and uncertainties

The combinatorics of track seeding has long been a computational bottleneck for triggering and offline computing in High Energy Physics (HEP), and remains so for the HL-LHC. Next-generation pixel sensors will be sufficiently fine-grained to determine angular information of the charged particle passing through from pixel-cluster properties. This detector technology immediately improves the situation for offline tracking, but any major improvements in physics reach are unrealized since they are dominated by lowest-level hardware trigger acceptance. We will demonstrate track angle and hit position prediction, including errors, using a mixture density network within a single layer of silicon as well as the progress towards and status of implementing the neural network in hardware on both FPGAs and ASICs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Surrogate-driven Variance-based Sensitivity Analysis of Thermal Storage Tanks in Integrated Energy Systems

Sensitivity analysis and uncertainty quantification are essential steps for enhancing the accuracy of computational models by identifying and mitigating uncertainties. This study focuses on these steps for the Thermal Energy Delivery System at Idaho National Laboratory, specifically targeting the thermocline tank. Using a Modelica/Dymola simulation model, the study perturbed various design parameters and boundary conditions, including shape factor, porosity, outlet temperature, inlet mass flow rate, and system pressure, to predict and quantify uncertainty in the tank’s ax- ial temperature. A dataset of over 1,000 simulations was generated, and surrogate models were developed using the pyMAISE (Michigan Artificial Intelligence Standard Environment) library, which is an Automatic Machine Learning library for nuclear engineering applications. The optimal model, a feedforward neural network with two hidden layers, achieved an R2 score above 0.99 and a mean absolute error below 1 Kelvin. Sensitivity analyses using Sobol indices and Fourier amplitude sensitivity testing methods on this surrogate model revealed that the inlet mass flow rate at initial timestamps and porosity significantly impacts predicted temperatures across all sensors and time steps.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗

Data-driven model for divertor plasma detachment prediction

We present a fast and accurate data-driven surrogate model for divertor plasma detachment prediction leveraging the latent feature space concept in machine learning research. Our approach involves constructing and training two neural networks: an autoencoder that finds a proper latent space representation (LSR) of plasma state by compressing the multi-modal diagnostic measurements and a forward model using multi-layer perception (MLP) that projects a set of plasma control parameters to its corresponding LSR. By combining the forward model and the decoder network from autoencoder, this new data-driven surrogate model is able to predict a consistent set of diagnostic measurements based on a few plasma control parameters. In order to ensure that the crucial detachment physics is correctly captured, highly efficient 1D UEDGE model is used to generate training and validation data in this study. The benchmark between the data-driven surrogate model and UEDGE simulations shows that our surrogate model is capable of providing accurate detachment prediction (usually within a few per cent relative error margin) but with at least four orders of magnitude speed-up, indicating that performance-wise, it has the potential to facilitate integrated tokamak design and plasma control. Comparing with the widely used two-point model and/or two-point model formatting, the new data-driven model features additional detachment front prediction and can be easily extended to incorporate richer physics. This study demonstrates that the complicated divertor and scrape-off-layer plasma state has a low-dimensional representation in latent space. Understanding plasma dynamics in latent space and utilising this knowledge could open a new path for plasma control in magnetic fusion energy research.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Star/galaxy separation at faint magnitudes: application to a simulated Dark Energy Survey

We address the problem of separating stars from galaxies in future large photometric surveys. We focus our analysis on simulations of the Dark Energy Survey (DES). In the first part of the paper, we derive the science requirements on star/galaxy separation, for measurement of the cosmological parameters with the gravitational weak lensing and large-scale structure probes. These requirements are dictated by the need to control both the statistical and systematic errors on the cosmological parameters, and by point spread function calibration. We formulate the requirements in terms of the completeness and purity provided by a given star/galaxy classifier. In order to achieve these requirements at faint magnitudes, we propose a new method for star/galaxy separation in the second part of the paper. We first use principal component analysis to outline the correlations between the objects parameters and extract from it the most relevant information. We then use the reduced set of parameters as input to an Artificial Neural Network. This multiparameter approach improves upon purely morphometric classifiers (such as the classifier implemented in SExtractor), especially at faint magnitudes: it increases the purity by up to 20 per cent for stars and by up to 12 per cent for galaxies, at i-magnitude fainter than 23.

79 ASTRONOMY AND ASTROPHYSICS↗

SU-D-BRD-01: An Automated Physics Weekly Chart Checking System Supporting ARIA

Purpose: A software tool was developed in this study to perform automatic weekly physics chart check on the patient data in ARIA. The tool accesses the electronic patient data directly from ARIA server and checks the accuracy of treatment deliveries, and generates reports which summarize the delivery history and highlight the errors. Methods: The tool has four modules. 1) The database interface is designed to directly access treatment delivery data from the ARIA database before reorganizing the data into the patient chart tree (PCT). 2) PCT is a core data structure designed to store and organize the data in logical hierarchies, and to be passed among functions. 3) The treatment data check module analyzes the organized data in PCT and stores the checking results into PCT. 4) Report generation module generates reports containing the treatment delivery summary, chart checking results and plots of daily treatment setup parameters (couch table positions, shifts of image guidance). The errors that are found by the tool are highlighted with colors. Results: The weekly check tool has been implemented in MATLAB and clinically tested at two major cancer centers. Javascript, cascading style sheets (CSS) and dynamic HTML were employed to create the user-interactive reports. It takes 0.06 second to search the delivery records of one beam with PCT and compare the delivery records with beam plan. The reports, saved in the HTML files on shared network folder, can be accessed by web browser on computers and mobile devices. Conclusion: The presented weekly check tool is useful to check the electronic patient treatment data in Varian ARIA system. It could be more efficient and reliable than the manually check by physicists. The work was partially supported by a research grant from Varian Medical System.

60 APPLIED LIFE SCIENCES↗

Star/galaxy separation at faint magnitudes: Application to a simulated Dark Energy Survey

We address the problem of separating stars from galaxies in future large photometric surveys. We focus our analysis on simulations of the Dark Energy Survey (DES). In the first part of the paper, we derive the science requirements on star/galaxy separation, for measurement of the cosmological parameters with the Gravitational Weak Lensing and Large Scale Structure probes. These requirements are dictated by the need to control both the statistical and systematic errors on the cosmological parameters, and by Point Spread Function calibration. We formulate the requirements in terms of the completeness and purity provided by a given star/galaxy classifier. In order to achieve these requirements at faint magnitudes, we propose a new method for star/galaxy separation in the second part of the paper. We first use Principal Component Analysis to outline the correlations between the objects parameters and extract from it the most relevant information. We then use the reduced set of parameters as input to an Artificial Neural Network. This multi-parameter approach improves upon purely morphometric classifiers (such as the classifier implemented in SExtractor), especially at faint magnitudes: it increases the purity by up to 20% for stars and by up to 12% for galaxies, at i-magnitude fainter than 23.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Predicting spring green-up across diverse North American grasslands

Vegetation phenology influences many ecosystem and climate processes, such as carbon uptake and energy and water cycles. Thus, understanding drivers of vegetation phenology is crucial for predicting current and future impacts of climate change on ecological systems. Existing models can accurately predict the date of spring green-up in temperate forests but tend to perform poorly in grassland systems. We hypothesize this is because most do not incorporate water availability, a primary limiting factor for grassland plants. In this study, we used long-term datasets of digital imagery from the PhenoCam Network of 43 diverse North American grassland sites (195 site-years) to test existing spring phenology models, as well as develop several new models that incorporate precipitation or soil moisture (53 models). As a result, most of the new models performed substantially better, with the best model requiring sufficient accumulated precipitation followed by warm temperatures to trigger spring onset (root mean square error, RMSE, between predicted and observed dates = 16.0 days). Importantly, the best model performed well across all grassland types using a single set of parameters, from temperate to arid grasslands. Since plants are adapted to their local climates, model performance was further improved when parameters were independently optimized for four separate climate regions (RMSE = 10.4 days). Therefore, both sufficient precipitation and temperature are required for grassland green-up, but optimal thresholds vary by region. Running the top model with projected climate data (representative concentration pathway 8.5) suggests that, depending on the climate region, spring onset will occur up to 12 days earlier within 100 years in temperature-limited sites, but the trend is unclear for precipitation-limited sites (3.5 ± 8.0 days later). This new phenology model improves our ability to understand and predict grassland dynamics, with implications for both current and future ecosystem processes related to carbon and water cycling.

54 ENVIRONMENTAL SCIENCES↗

Adaptive Image-Guided Radiotherapy (IGRT) Eliminates the Risk of Biochemical Failure Caused by the Bias of Rectal Distension in Prostate Cancer Treatment Planning: Clinical Evidence

Purpose: Rectal distension has been shown to decrease the probability of biochemical control. Adaptive image-guided radiotherapy (IGRT) corrects for target position and volume variations, reducing the risk of biochemical failure while yielding acceptable rates of gastrointestinal (GI)/genitourinary (GU) toxicities. Methods and Materials: Between 1998 and 2006, 962 patients were treated with computed tomography (CT)-based offline adaptive IGRT. Patients were stratified into low (n = 400) vs. intermediate/high (n = 562) National Comprehensive Cancer Network (NCCN) risk groups. Target motion was assessed with daily CT during the first week. Electronic portal imaging device (EPID) was used to measure daily setup error. Patient-specific confidence-limited planning target volumes (cl-PTV) were then constructed, reducing the standard PTV and compensating for geometric variation of the target and setup errors. Rectal volume (RV), cross-sectional area (CSA), and rectal volume from the seminal vesicles to the inferior prostate (SVP) were assessed on the planning CT. The impact of these volumetric parameters on 5-year biochemical control (BC) and chronic Grades {>=}2 and 3 GU and GI toxicity were examined. Results: Median follow-up was 5.5 years. Median minimum dose covering cl-PTV was 75.6 Gy. Median values for RV, CSA, and SVP were 82.8 cm{sup 3}, 5.6 cm{sup 2}, and 53.3 cm{sup 3}, respectively. The 5-year BC was 89% for the entire group: 96% for low risk and 83% for intermediate/high risk (p < 0.001). No statistically significant differences in BC were seen with stratification by RV, CSA, and SVP in quartiles. Maximum chronic Grades {>=}2 and 3 GI toxicities were 21.2% and 2.9%, respectively. Respective values for GU toxicities were 15.5% and 4.3%. No differences in GI or GU toxicities were noted when patients were stratified by RV. Conclusions: Incorporation of adaptive IGRT reduces the risk of geometric miss and results in excellent biochemical control that is independent of rectal volume/distension while maintaining very low rates of chronic GI toxicity.

62 RADIOLOGY AND NUCLEAR MEDICINE↗