Search NASA⌕ Search

SEARCH · Search NASA

Results for “Guided Dropout”

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.

7 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Power Grid Contingency Analysis with Machine Learning: A Brief Survey and Prospects

We briefly review previous applications of machine learning (ML) in power grid analyses and introduce our ongoing effort toward developing a generative-adversarial (GA) model for fast and reliable grid contingency analyses. According to our review, the persisting limitation of traditional ML techniques in grid analyses is the need for an exhaustive amount of training data for model generalization and accurate predictions. GA models overcome this limitation by first learning true data distribution from a small training set, from which new samples assimilating true data are generated with some variations. Subsequently, GA models can transfer learn or super-generalize with increased accuracy, that is, accurately predict n - (k + 2) contingencies from a small n - k training set and generated n - (k + 1) data. The joint effort between Idaho National Lab and Florida State University strives to develop a zero-shot and deep learning-based contingency analysis tool, named Smart Contingency Analysis Neural Network (SCANN), by leveraging the aforementioned advantages of GA models. The basic architecture of SCANN stems from the Latent Encoding of Atypical Perturbations network combined with an adversarial network, and it is designed to generate imbalanced power flow data from learned true data distributions for prediction purposes. Here we also introduce the abstract concept of resilience-chaos plots, a new resilience characterization tool proposed to complement SCANN by aiding in the assessment of large amounts of high-order contingency predictions.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Multiplex Editing of the Nucleoredoxin1 Tandem Array in Poplar: From Small Indels to Translocations and Complex Inversions

The CRISPR-Cas9 system has been deployed for precision mutagenesis in an ever-growing number of species, including agricultural crops and forest trees. Its application to closely linked genes with extremely high sequence similarities has been less explored. In this study, we used CRISPR-Cas9 to mutagenize a tandem array of seven Nucleoredoxin1 (NRX1) genes spanning ~100 kb in Populus tremula × Populus alba. We demonstrated efficient multiplex editing with one single guide RNA in 42 transgenic lines. The mutation profiles ranged from small insertions and deletions and local deletions in individual genes to large genomic dropouts and rearrangements spanning tandem genes. We also detected complex rearrangements including translocations and inversions resulting from multiple cleavage and repair events. Target capture sequencing was instrumental for unbiased assessments of repair outcomes to reconstruct unusual mutant alleles. The work highlights the power of CRISPR-Cas9 for multiplex editing of tandemly duplicated genes to generate diverse mutants with structural and copy number variations to aid future functional characterization.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Multiplex knockout of trichome-regulating MYB duplicates in hybrid poplar using a single gRNA

As the focus for CRISPR/Cas-edited plants moves from proof-of-concept to real-world applications, precise gene manipulation will increasingly require concurrent multiplex editing for polygenic traits. A common approach for editing across multiple sites is to design one guide RNA (gRNA) per target; however, this complicates construct assembly and increases the possibility of off-target mutations. In this study, we utilized one gRNA to target MYB186, a known positive trichome regulator, as well as its paralogs MYB138 and MYB38 at a consensus site for mutagenesis in hybrid poplar (Populus tremula × P. alba INRA 717-1B4). Unexpected duplications of MYB186 and MYB138 resulted in eight alleles for the three targeted genes in the hybrid poplar. Deep sequencing and polymerase chain reaction analyses confirmed editing across all eight targets in nearly all of the resultant glabrous mutants, ranging from small indels to large genomic dropouts, with no off-target activity detected at four potential sites. This highlights the effectiveness of a single gRNA targeting conserved exonic regions for multiplex editing. Additionally, cuticular wax and whole-leaf analyses showed a complete absence of triterpenes in the trichomeless mutants, hinting at a previously undescribed role for the nonglandular trichomes of poplar.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS2005

Fuzzy Logic Particle Tracking

A new all-electronic Particle Image Velocimetry technique that can efficiently map high speed gas flows has been developed in-house at the NASA Lewis Research Center. Particle Image Velocimetry is an optical technique for measuring the instantaneous two component velocity field across a planar region of a seeded flow field. A pulsed laser light sheet is used to illuminate the seed particles entrained in the flow field at two instances in time. One or more charged coupled device (CCD) cameras can be used to record the instantaneous positions of particles. Using the time between light sheet pulses and determining either the individual particle displacements or the average displacement of particles over a small subregion of the recorded image enables the calculation of the fluid velocity. Fuzzy logic minimizes the required operator intervention in identifying particles and computing velocity. Using two cameras that have the same view of the illumination plane yields two single exposure image frames. Two competing techniques that yield unambiguous velocity vector direction information have been widely used for reducing the single-exposure, multiple image frame data: (1) cross-correlation and (2) particle tracking. Correlation techniques yield averaged velocity estimates over subregions of the flow, whereas particle tracking techniques give individual particle velocity estimates. For the correlation technique, the correlation peak corresponding to the average displacement of particles across the subregion must be identified. Noise on the images and particle dropout result in misidentification of the true correlation peak. The subsequent velocity vector maps contain spurious vectors where the displacement peaks have been improperly identified. Typically these spurious vectors are replaced by a weighted average of the neighboring vectors, thereby decreasing the independence of the measurements. In this work, fuzzy logic techniques are used to determine the true correlation displacement peak even when it is not the maximum peak, hence maximizing the information recovery from the correlation operation, maintaining the number of independent measurements, and minimizing the number of spurious velocity vectors. Correlation peaks are correctly identified in both high and low seed density cases. The correlation velocity vector map can then be used as a guide for the particle-tracking operation. Again fuzzy logic techniques are used, this time to identify the correct particle image pairings between exposures to determine particle displacements, and thus the velocity. Combining these two techniques makes use of the higher spatial resolution available from the particle tracking. Particle tracking alone may not be possible in the high seed density images typically required for achieving good results from the correlation technique. This two-staged velocimetric technique can measure particle velocities with high spatial resolution over a broad range of seeding densities.

Source record↗
NASA NTRS1995

Fuzzy inference enhanced information recovery from digital PIV using cross-correlation combined with particle tracking

Particle Image Velocimetry provides a means of measuring the instantaneous 2-component velocity field across a planar region of a seeded flowfield. In this work only two camera, single exposure images are considered where both cameras have the same view of the illumination plane. Two competing techniques which yield unambiguous velocity vector direction information have been widely used for reducing the single exposure, multiple image data: cross-correlation and particle tracking. Correlation techniques yield averaged velocity estimates over subregions of the flow, whereas particle tracking techniques give individual particle velocity estimates. The correlation technique requires identification of the correlation peak on the correlation plane corresponding to the average displacement of particles across the subregion. Noise on the images and particle dropout contribute to spurious peaks on the correlation plane, leading to misidentification of the true correlation peak. The subsequent velocity vector maps contain spurious vectors where the displacement peaks have been improperly identified. Typically these spurious vectors are replaced by a weighted average of the neighboring vectors, thereby decreasing the independence of the measurements. In this work fuzzy logic techniques are used to determine the true correlation displacement peak even when it is not the maximum peak on the correlation plane, hence maximizing the information recovery from the correlation operation, maintaining the number of independent measurements and minimizing the number of spurious velocity vectors. Correlation peaks are correctly identified in both high and low seed density cases. The correlation velocity vector map can then be used as a guide for the particle tracking operation. Again fuzzy logic techniques are used, this time to identify the correct particle image pairings between exposures to determine particle displacements, and thus velocity. The advantage of this technique is the improved spatial resolution which is available from the particle tracking operation. Particle tracking alone may not be possible in the high seed density images typically required for achieving good results from the correlation technique. This two staged approach offers a velocimetric technique capable of measuring particle velocities with high spatial resolution over a broad range of seeding densities.

Wernet, Mark P.↗
DOE OSTI2022

Deep structural clustering for single-cell RNA-seq data jointly through autoencoder and graph neural network

Abstract Single-cell RNA sequencing (scRNA-seq) permits researchers to study the complex mechanisms of cell heterogeneity and diversity. Unsupervised clustering is of central importance for the analysis of the scRNA-seq data, as it can be used to identify putative cell types. However, due to noise impacts, high dimensionality and pervasive dropout events, clustering analysis of scRNA-seq data remains a computational challenge. Here, we propose a new deep structural clustering method for scRNA-seq data, named scDSC, which integrate the structural information into deep clustering of single cells. The proposed scDSC consists of a Zero-Inflated Negative Binomial (ZINB) model-based autoencoder, a graph neural network (GNN) module and a mutual-supervised module. To learn the data representation from the sparse and zero-inflated scRNA-seq data, we add a ZINB model to the basic autoencoder. The GNN module is introduced to capture the structural information among cells. By joining the ZINB-based autoencoder with the GNN module, the model transfers the data representation learned by autoencoder to the corresponding GNN layer. Furthermore, we adopt a mutual supervised strategy to unify these two different deep neural architectures and to guide the clustering task. Extensive experimental results on six real scRNA-seq datasets demonstrate that scDSC outperforms state-of-the-art methods in terms of clustering accuracy and scalability. Our method scDSC is implemented in Python using the Pytorch machine-learning library, and it is freely available at https://github.com/DHUDBlab/scDSC.

Gan, Yanglan↗
DOE OSTI2021

Machine Learning Assisted Safety Modeling and Analysis of Advanced Reactors

With the advances in computational power and numerical methods, analysts can now rely on first-principle simulations to predict ultra-fine details in a variety of applications. Advances in machine learning (ML) have produced algorithms that can now learn high-level abstractions via hierarchical models. This project aims to leverage advances in ML techniques and the available high-resolution simulation data to develop a novel modeling and simulation (M\&S) methodology for reactor safety analysis. While application-agnostic ML techniques are available, complex physics constraints need to be incorporated into ML techniques to build ML-based closures for computationally efficient predictive simulations. This project intends to develop a physics-guided data-driven multi-scale methodology for M\&S of advanced reactors. The project focuses on thermal fluid (T/F) phenomena, which play major roles in advanced reactor safety. Specifically, we propose a data-driven coarse-mesh turbulence model based on local flow features for the transient analysis of thermal mixing and stratification in a sodium-cooled fast reactor (SFR). The model has a coarse-mesh setup to ensure computational efficiency, while it is trained by fine-mesh computational fluid dynamics (CFD) data with Reynolds-averaged Navier-Stokes (RANS) turbulence model to ensure accuracy. Three different neural networks are developed and tested for loss-of-flow transients in the hot pool of SFR, i.e. the densely connected convolutional neural network (DCNN), long-short-term-memory network based on proper orthogonal decomposition (POD-LSTM), and the DCNN informed by LSTM (DCNN-LSTM). The performances of these three neural networks are evaluated based on baseline models. The DCNN-LSTM model has been chosen for further hyperparameter optimization. Furthermore, based on a simplified two-dimensional case, uncertainty quantification (UQ) of the developed ML-based closure are investigated with three methods, i.e. Monte Carlo dropout, deep ensemble, and Bayesian neural network. The developed ML-based turbulent viscosity closure relation based on deep ensemble is then integrated into the system analysis module SAM and serves as a term in the conservation equations. Such a SAM-ML based procedure guarantees that the obtained results are consistent with the physical constraints of the thermal-fluid system. The SAM-ML simulation on the same loss-of-flow transient showed comparable accuracy with the CFD simulation but with a much coarser mesh setup. Last but not least, the ML-based closure improvement with the support of higher-fidelity data from large eddy simulation (LES) is discussed. As a first step towards this direction, a baseline LES simulation is performed to obtain comparable data with RANS results. Based on the early results, future investigation on further improving the ML-based closure is discussed. We believe the developed approach that combines scientific machine learning with nuclear system analysis code can benefit the advanced reactor community as more accurate safety analyses will better characterize reactor safety margins and reduce licensing efforts.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗