Search NASA⌕ Search

SEARCH · Search NASA

Results for “gradient based optimization”

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 757 records · Page 42
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Quench process modeling & simulation in the heat-treatment of critical aerospace components

To manufacture light-weight, advanced metal alloy components for gas turbine engines, quench heat-treatment processes are typically used. By quenching the component from elevated tempera-tures, the alloy sometimes undergoes a solid-state phase transformation which produces special microstructures with the required, enhanced mechanical properties. However, the quenching can also lead to cracks forming in the component. Addressing the quench cracking problems adds a significant burden to the cost, schedule, and energy demand of manufacture. Currently, optimizing the quench process to mitigate or avoid the cracking is performed largely by trial-and-error, relying heavily on costly experimental (thermocouple) trials to understand the local thermal gradients which cause the cracks to form. In this first part (Phase 1) of the work, high-performance computing is employed to establish the ability of modern CFD (computational fluid dynamics) to alleviate or wholly replace the experimental quenching trials by virtual testing. A baseline CFD model is defined and its accuracy established to be comparable to (and which usually exceeds) the accuracy of existing HTC (heat-transfer correlation) based simulation methods of quenching. As a first-principles based approach, “calibration” of the Baseline CFD model is independent of the quench process itself, but instead relies on the accuracy of the underlying (modeled), generic two-phase fluid processes which cannot be currently resolved by CFD for large, industrial-scale cases. A novel, high-fidelity DNS capability has been developed and verified to examine and further improve upon the mean-field closure submodels on which the Baseline CFD approach is based.

36 MATERIALS SCIENCE↗
NASA NTRS1995

Optimization on a Network-based Parallel Computer System for Supersonic Laminar Wing Design

A set of Computational Fluid Dynamics (CFD) routines and flow transition prediction tools are integrated into a network based parallel numerical optimization routine. Through this optimization routine, the design of a 2-D airfoil and an infinitely swept wing will be studied in order to advance the design cycle capability of supersonic laminar flow wings. The goal of advancing supersonic laminar flow wing design is achieved by wisely choosing the design variables used in the optimization routine. The design variables are represented by the theory of Fourier series and potential theory. These theories, combined with the parallel CFD flow routines and flow transition prediction tools, provide a design space for a global optimal point to be searched. Finally, the parallel optimization routine enables gradient evaluations to be performed in a fast and parallel fashion.

Garcia, Joseph A.↗
DOE OSTI2022

Spatial Graph Attention and Curiosity-driven Policy for Antiviral Drug Discovery

We developed Distilled Graph Attention Policy Network (DGAPN), a reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. The framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention (sGAT) mechanism that leverages self-attention over both node and edge attributes as well as encoding the spatial structure --- this capability is of considerable interest in synthetic biology and drug discovery. An attentional policy network is introduced to learn the decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with stability. Exploration is driven by the stochasticity of the action space design and the innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while reducing the complexity of paths to chemical synthesis.

Wu, Yulun↗
DOE OSTI2022

SPATIAL GRAPH ATTENTION AND CURIOSITY-DRIVEN POLICY FOR ANTIVIRAL DRUG DISCOVERY

We developed Distilled Graph Attention Policy Network (DGAPN), a reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. The framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention (sGAT) mechanism that leverages self-attention over both node and edge attributes as well as encoding the spatial structure - this capability is of considerable interest in synthetic biology and drug discovery. An attentional policy network is introduced to learn the decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with stability. Exploration is driven by the stochasticity of the action space design and the innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while reducing the complexity of paths to chemical synthesis.

Wu, Y↗
DOE OSTI2025

Genomic insights into local adaptation and migration success in reintroduced Coho Salmon of the Wenatchee River basin

ABSTRACT Objective Reintroduction of salmonids into regions where they have been extirpated is a common conservation strategy that is often implemented through natural recolonization, translocation of natural populations, or hatchery-based programs. Locally adapting to specific environmental conditions is critical for long-term population viability, particularly for species like Coho Salmon Oncorhynchus kisutch, which face diverse selective pressures during their migration. This study focused on the mid-Columbia River Coho Salmon reintroduction program managed by Yakama Nation Fisheries, which has successfully reintroduced Coho Salmon into the Wenatchee and Methow River basins, Washington. Notably, these populations have adapted to the longer migration route than those in the founding stock, with selection favoring individuals with an earlier arrival time and that can navigate a 15-km, high-gradient canyon to reach optimal spawning grounds. The objectives of this study were to investigate whether specific genomic regions are under selection for traits associated with return location and timing in Coho Salmon. Methods Low-coverage whole-genome resequencing data were used to screen for genomic regions associated with the phenotypes of interest. Results A weak polygenic signal in female Coho Salmon was found to be associated with return group, with a subset of candidate adaptive regions occurring across eight chromosomes. Conclusions These findings provide insights into the genomic mechanisms underlying local adaptation in reintroduced salmon populations and inform broodstock selection strategies aimed at promoting natural production and long-term population sustainability.

Horn, Rebekah L.↗
DOE OSTI2026

Accelerated creep profiling: a high-throughput thermal- and stress-gradient approach applied to additive and wrought stainless steel

Here, this study investigates logarithmic creep behavior at temperatures near room temperature in wrought and additively manufactured (AM) 316 L austenitic stainless steel. A novel high-throughput (HiTp) methodology employed thermal gradients to obtain stress- and temperature-dependence of creep behavior. The findings, based on 104 two-week creep tests, reveal significant differences in creep response between wrought and AM materials. Compared to the wrought material, the AM material exhibited a higher activation energy and a lower activation volume. These differences were attributed to the materials’ distinct microstructural features, e.g., dislocation cell structures and precipitate distributions, which alter the susceptibility to dislocation glide creep, particularly at stresses below the yield stress. The HiTp methodology provides both an accelerated determination of low-temperature creep mechanisms, offering critical data for the design and optimization of alloys in near-ambient, high stress applications.

316 L Stainless steel↗
DOE OSTI2025

A Two-Stage Quantum Reinforcement Learning Method for Multi-Objective Transmission Switching

Multi-objective transmission switching (MO-TS) problems involve the strategic reconfiguration of network topology to simultaneously optimize multiple objectives. As the system scale increases, finding feasible solutions becomes increasingly challenging due to the problem's nonlinearity and high computational complexity. To address these challenges, this paper proposes a two-stage quantum reinforcement learning method that leverages potential quantum advantages for MO-TS. In the first stage, candidate switching lines are identified using a graph-theoretical approach to reduce the problem's dimensionality. The second stage introduces a quantum-classical reinforcement learning framework, where a learnable measurement-based CNN-ResVQC architecture is developed to effectively reduce the input dimension for quantum processing, mitigate vanishing gradients, and enhance trainability while improving the quantum circuit's flexibility in modeling complex decision policies for MO-TS. Numerical studies on IEEE 14-bus, 57-bus, and 118-bus systems demonstrate that the proposed algorithm achieves superior training stability and faster convergence with approximately 1% of the network parameters required by classical algorithms, highlighting its effectiveness, efficiency, and scalability. Furthermore, the practicality is validated through its stable convergence under three common quantum noise channels.

99 GENERAL AND MISCELLANEOUS↗
NASA NTRSDate not supplied

Simulations of a Turbulent Flow Subjected to Favorable and Adverse Pressure Gradients

This paper reports the results from a direct numerical simulation of an initially turbulent boundary layer passing over a wall-mounted “speed bump” geometry. The speed bump, represented in the form of a Gaussian distribution profile, generates a favorable pressure gradient region over the upstream half of the geometry, followed by an adverse pressure gradient over the downstream half. The boundary layer approaching the bump undergoes strong acceleration in the favorable pressure gradient region before experiencing incipient or very weak separation within the adverse pressure gradient region. These types of flows have proven to be particularly challenging to predict using lower-fidelity simulation tools based on various turbulence modeling approaches and warrant the use of the highest-fidelity simulation techniques. Simulation results are utilized to examine the key phenomena present in the flowfield, such as relaminarization/stabilization in the strong acceleration region succeeded by retransition to turbulence near the onset of adverse pressure gradient, incipient/weak separation, and development of internal layers where the sense of streamwise pressure gradient changes at the foot, apex and tail of the bump. The present direct numerical simulation is performed using a flow solver developed exclusively for graphics processing units, which is found to provide a significant speedup compared to an earlier solver optimized for central processing unit architectures.

Ali Uzun↗
DOE OSTI2024

Super Resolving Unrolled Neural Networks for Remote Sensing

In remote sensing systems, the capabilities of the system are constrained by the complex interactions between size, weight, and power (SWAP) of potential designs. In electro-optical (EO) systems, examples of these critical parameters include the system’s sensitivity and resolution. Those parameters can be increased by ever larger optical apertures and focal planes but at the cost of more SWAP. Multi-image super resolution (MISR) techniques allow resolution to be enhanced via computation rather than more sophisticated optical hardware. These algorithms combine multiple images together into a single, higher resolution image, trading temporal resolution and computation for spatial resolution. Fielded MISR techniques, such as Drizzle, can require several hundred images to create a single super resolved image, implying reduced temporal resolution, increased data acquisition load, and limiting mission applications. Iterative techniques, such as model-based image reconstruction and compressive sensing, have been shown to create super resolved images using fewer images than Drizzle. They do this by posing an optimization problem that balances accuracy between a highly accurate physical model and an image model. In the case of super resolution, the physical model is defined by the relation between low resolution input images and the desired high resolution output image. The image model encodes some assumptions about the super resolved image. These assumptions are meant to suppress reconstruction artifacts that arise due to deterministic physical model error, stochastic measurement noise, and potential undersampling. In practice, the performance of iterative methods are limited by imaging models compatible with optimization. Deep learning-based methods can effectively learn image models of arbitrary complexity, but lack the theoretical explainability and robustness of iterative techniques. Consensus equilibrium (CE) generalizes the iterative techniques beyond optimization, enabling blackbox algorithms such as traditional and neural image denoisers to be used as the image model. CE-based approaches retain much of the explainability and robustness of iterative techniques while allowing the expressiveness of machine learning image models to be used. Additionally, by unrolling iterations of CE with an embedded image denoiser, the image denoiser can be further trained and specialized to the specific application with potentially higher quality reconstructions. Under this project, we demonstrated the feasibility of training an unrolled neural network based upon CE. While we didn’t train one, we showed that the CE process is differentiable and its gradient can be tractably computed. We also explored the usage of a variants of CE akin to generative neural works. Most importantly, we applied the CE framework to a number of problems including non-blind deconvolution, upsampling, single-image super resolution, MISR, event-based sensing, and saturated deconvolution. Our MISR prototype creates high quality reconstructions with an order of magnitude fewer images than previous approaches and, critically, produces these reconstructions fast enough for practical usage.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Spatial Graph Attention and Curiosity-driven Policy for Antiviral Drug Discovery

We developed Distilled Graph Attention Policy Networks (DGAPNs), a curiosity-driven reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. The framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention Network (sGAT) that leverages self-attention over both node and edge attributes as well as encoding spatial structure -- this capability is of considerable interest in areas such as molecular and synthetic biology and drug discovery. An attentional policy network is then introduced to learn decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with enhanced stability. Exploration is efficiently encouraged by incorporating innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while increasing the diversity of proposed molecules and reducing the complexity of paths to chemical synthesis.

Wu, Yulun↗
NASA NTRS1976

An iterative procedure for obtaining maximum-likelihood estimates of the parameters for a mixture of normal distributions, 2

The problem of obtaining numerically maximum likelihood estimates of the parameters for a mixture of normal distributions is addressed. In recent literature, a certain successive approximations procedure, based on the likelihood equations, is shown empirically to be effective in numerically approximating such maximum-likelihood estimates; however, the reliability of this procedure was not established theoretically. Here, a general iterative procedure is introduced, of the generalized steepest-ascent (deflected-gradient) type, which is just the procedure known in the literature when the step-size is taken to be 1. With probability 1 as the sample size grows large, it is shown that this procedure converges locally to the strongly consistent maximum-likelihood estimate whenever the step-size lies between 0 and 2. The step-size which yields optimal local convergence rates for large samples is determined in a sense by the separation of the component normal densities and is bounded below by a number between 1 and 2.

Peters, B. C., Jr.↗
NASA NTRS2026

Incremental Learning for Passive Microwave Precipitation Retrievals using Advanced Technology Microwave Sounder

Spaceborne passive microwave (PMW) radiometry is central to global precipitation monitoring, yet retrieval uncertainties remain substantial, particularly for cross-track sounders whose variable footprints and channel configurations are optimized for atmospheric temperature and moisture profiling rather than precipitation. Consequently, existing operational products often exhibit angular-dependent biases, limited effective swath utilization, unrealistic rainfall probability distributions, and systematic misclassification of precipitation phase. These limitations are further compounded by the scarcity of globally accurate and representative precipitation observations, as training data from the Dual-frequency Precipitation Radar (DPR) and the Cloud Profiling Radar (CPR) are spatially sparse, lack uniform global coverage, and exhibit heterogeneous error characteristics across precipitation regimes. To address these challenges, this study presents a supervised retrieval algorithm that incrementally trains an ensemble of extreme gradient-boosted decision trees by augmenting base learners with pre-training on reanalysis data and post-training on coincident DPR and CPR observations matched with the Advanced Technology Microwave Sounder (ATMS). By transferring prior information from reanalysis to posterior constraints from radar observations and adopting a sequential detection–estimation strategy for precipitation phase and rate retrieval, the proposed approach yields retrievals across the full ATMS swath that are largely free from persistent deficiencies in current Global Precipitation Measurement (GPM) passive microwave operational products. In particular, the method resolves bimodal artifacts in rainfall retrievals and mitigates systematic high-latitude snowfall biases, including overestimation across the Arctic and underestimation across the Antarctic. Validation against independent Multi-Radar Multi-Sensor (MRMS) data over the Contiguous United States (CONUS) further demonstrates improved performance in precipitation phase detection and rate estimation relative to both reanalysis and current GPM PMW products.

Mahyar Garshasbi↗
DOE OSTI2021

Spatial Graph Attention and Curiosity-driven Policy for Antiviral Drug Discovery

We developed Distilled Graph Attention Policy Networks (DGAPNs), a curiosity-driven reinforcement learning model to generate novel graph-structured chemical representations that optimize user-defined objectives by efficiently navigating a physically constrained domain. Here, the framework is examined on the task of generating molecules that are designed to bind, noncovalently, to functional sites of SARS-CoV-2 proteins. We present a spatial Graph Attention Network (sGAT) that leverages self-attention over both node and edge attributes as well as encoding spatial structure -- this capability is of considerable interest in areas such as molecular and synthetic biology and drug discovery. An attentional policy network is then introduced to learn decision rules for a dynamic, fragment-based chemical environment, and state-of-the-art policy gradient techniques are employed to train the network with enhanced stability. Exploration is efficiently encouraged by incorporating innovation reward bonuses learned and proposed by random network distillation. In experiments, our framework achieved outstanding results compared to state-of-the-art algorithms, while increasing the diversity of proposed molecules and reducing the complexity of paths to chemical synthesis.

Wu, Yulun↗
DOE OSTI2019

Towards optimized viral metagenomes for double-stranded and single-stranded DNA viruses from challenging soils

Soils impact global carbon cycling and their resident microbes are critical to their biogeochemical processing and ecosystem outputs. Based on studies in marine systems, viruses infecting soil microbes likely modulate host activities via mortality, horizontal gene transfer, and metabolic control. However, their roles remain largely unexplored due to technical challenges with separating, isolating, and extracting DNA from viruses in soils. Some of these challenges have been overcome by using whole genome amplification methods and while these have allowed insights into the identities of soil viruses and their genomes, their inherit biases have prevented meaningful ecological interpretations. Here we experimentally optimized steps for generating quantitatively-amplified viral metagenomes to better capture both ssDNA and dsDNA viruses across three distinct soil habitats along a permafrost thaw gradient. First, we assessed differing DNA extraction methods (PowerSoil, Wizard mini columns, and cetyl trimethylammonium bromide) for quantity and quality of viral DNA. This established PowerSoil as best for yield and quality of DNA from our samples, though ∼1/3 of the viral populations captured by each extraction kit were unique, suggesting appreciable differential biases among DNA extraction kits. Second, we evaluated the impact of purifying viral particles after resuspension (by cesium chloride gradients; CsCl) and of viral lysis method (heat vs bead-beating) on the resultant viromes. DNA yields after CsCl particle-purification were largely non-detectable, while unpurified samples yielded 1–2-fold more DNA after lysis by heat than by bead-beating. Virome quality was assessed by the number and size of metagenome-assembled viral contigs, which showed no increase after CsCl-purification, but did from heat lysis relative to bead-beating. We also evaluated sample preparation protocols for ssDNA virus recovery. In both CsCl-purified and non-purified samples, ssDNA viruses were successfully recovered by using the Accel-NGS 1S Plus Library Kit. While ssDNA viruses were identified in all three soil types, none were identified in the samples that used bead-beating, suggesting this lysis method may impact recovery. Further, 13 ssDNA vOTUs were identified compared to 582 dsDNA vOTUs, and the ssDNA vOTUs only accounted for ∼4% of the assembled reads, implying dsDNA viruses were dominant in these samples. This optimized approach was combined with the previously published viral resuspension protocol into a sample-to-virome protocol for soils now available at protocols.io , where community feedback creates ‘living’ protocols. This collective approach will be particularly valuable given the high physicochemical variability of soils, which will may require considerable soil type-specific optimization. This optimized protocol provides a starting place for developing quantitatively-amplified viromic datasets and will help enable viral ecogenomic studies on organic-rich soils.

dsDNA viruses↗
DOE OSTI2024

Microscale Thermophoresis (MST) as a Tool to Study Binding Interactions of Oxygen-Sensitive Biohybrids

Microscale thermophoresis (MST) is a technique used to measure the strength of molecular interactions. MST is a thermophoretic-based technique that monitors the change in fluorescence associated with the movement of fluorescent-labeled molecules in response to a temperature gradient triggered by an IR LASER. MST has advantages over other approaches for examining molecular interactions, such as isothermal titration calorimetry, nuclear magnetic resonance, biolayer interferometry, and surface plasmon resonance, requiring a small sample size that does not need to be immobilized and a high-sensitivity fluorescence detection. In addition, since the approach involves the loading of samples into capillaries that can be easily sealed, it can be adapted to analyze oxygen-sensitive samples. In this Bio-protocol, we describe the troubleshooting and optimization we have done to enable the use of MST to examine protein–protein interactions, protein–ligand interactions, and protein–nanocrystal interactions. The salient elements in the developed procedures include 1) loading and sealing capabilities in an anaerobic chamber for analysis using a NanoTemper MST located on the benchtop in air, 2) identification of the optimal reducing agents compatible with data acquisition with effective protection against trace oxygen, and 3) the optimization of data acquisition and analysis procedures. The procedures lay the groundwork to define the determinants of molecular interactions in these technically demanding systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1992

Study of Fluid Experiment System (FES)/CAST/Holographic Ground System (HGS)

The use of holographic and schlieren optical techniques for studying the concentration gradients in solidification processes has been used by several investigators over the years. The HGS facility at MSFC has been primary resource in researching this capability. Consequently, scientific personnel have been able to utilize these techniques in both ground based research and in space experiments. An important event in the scientific utilization of the HGS facilities was the TGS Crystal Growth and the casting and solidification technology (CAST) experiments that were flown on the International Microgravity Laboratory (IML) mission in March of this year. The preparation and processing of these space observations are the primary experiments reported in this work. This project provides some ground-based studies to optimize on the holographic techniques used to acquire information about the crystal growth processes flown on IML. Since the ground-based studies will be compared with the space-based experimental results, it is necessary to conduct sufficient ground based studies to best determine how the experiment worked in space. The current capabilities in computer based systems for image processing and numerical computation have certainly assisted in those efforts. As anticipated, this study has certainly shown that these advanced computing capabilities are helpful in the data analysis of such experiments.

Workman, Gary L.↗
DOE OSTI2020

DPM: A deep learning PDE augmentation method with application to large-eddy simulation

A framework is introduced that leverages known physics to reduce overfitting in machine learning for scientific applications. The partial differential equation (PDE) that expresses the physics is augmented with a neural network that uses available data to learn a description of the corresponding unknown or unrepresented physics. Training within this combined system corrects for missing, unknown, or erroneously represented physics, including discretization errors associated with the PDE's numerical solution. For optimization of the network within the PDE, an adjoint PDE is solved to provide high-dimensional gradients, and a stochastic adjoint method (SAM) further accelerates training. Additionally, the approach is demonstrated for large-eddy simulation (LES) of turbulence. High-fidelity direct numerical simulations (DNS) of decaying isotropic turbulence provide the training data used to learn sub-filter-scale closures for the filtered Navier–Stokes equations. Out-of-sample comparisons show that the deep learning PDE method outperforms widely-used models, even for filter sizes so large that they become qualitatively incorrect. It also significantly outperforms the same neural network when a priori trained based on simple data mismatch, not accounting for the full PDE. Measures of discretization errors, which are well-known to be consequential in LES, point to the importance of the unified training formulation's design, which without modification corrects for them. For comparable accuracy, simulation runtime is significantly reduced. A relaxation of the typical discrete enforcement of the divergence-free constraint in the solver is also successful, instead allowing the DPM to approximately enforce incompressibility physics. Since the training loss function is not restricted to correspond directly to the closure to be learned, training can incorporate diverse data, including experimental data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS1978

An iterative procedure for obtaining maximum-likelihood estimates of the parameters for a mixture of normal distributions

This paper addresses the problem of obtaining numerically maximum-likelihood estimates of the parameters for a mixture of normal distributions. In recent literature, a certain successive-approximations procedure, based on the likelihood equations, was shown empirically to be effective in numerically approximating such maximum-likelihood estimates; however, the reliability of this procedure was not established theoretically. Here, we introduce a general iterative procedure, of the generalized steepest-ascent (deflected-gradient) type, which is just the procedure known in the literature when the step-size is taken to be 1. We show that, with probability 1 as the sample size grows large, this procedure converges locally to the strongly consistent maximum-likelihood estimate whenever the step-size lies between 0 and 2. We also show that the step-size which yields optimal local convergence rates for large samples is determined in a sense by the 'separation' of the component normal densities and is bounded below by a number between 1 and 2.

Peters, B. C., Jr.↗