SEARCH · Engineering Papers

Results for “Domain adaptation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 559 records · Page 31

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Gaming the beamlines—employing reinforcement learning to maximize scientific outcomes at large-scale user facilities

Abstract Beamline experiments at central facilities are increasingly demanding of remote, high-throughput, and adaptive operation conditions. To accommodate such needs, new approaches must be developed that enable on-the-fly decision making for data intensive challenges. Reinforcement learning (RL) is a domain of AI that holds the potential to enable autonomous operations in a feedback loop between beamline experiments and trained agents. Here, we outline the advanced data acquisition and control software of the Bluesky suite, and demonstrate its functionality with a canonical RL problem: cartpole. We then extend these methods to efficient use of beamline resources by using RL to develop an optimal measurement strategy for samples with different scattering characteristics. The RL agents converge on the empirically optimal policy when under-constrained with time. When resource limited, the agents outperform a naive or sequential measurement strategy, often by a factor of 100%. We interface these methods directly with the data storage and provenance technologies at the National Synchrotron Light Source II, thus demonstrating the potential for RL to increase the scientific output of beamlines, and layout the framework for how to achieve this impact.

36 MATERIALS SCIENCE↗

NASA NTRS2002

Development of MLPG and LBIE Methods for Nonlinear Problems of Fracture

The truly meshless local Petrov-Galerkin (MLPG) method holds a great promise in solving boundary value problems, using a local symmetric weak form as a natural approach. In the present paper, in the context of MLPG and the meshless interpolation of a moving least squares (MLS) type, a method which uses primary and secondary nodes in the domain and on the global boundary is introduced, in order to improve the accuracy of solution. The secondary nodes can be placed at any location where one needs to obtain a better resolution. The sub-domains for the shape functions in the MLS approximation are defined only from the primary nodes, and the secondary nodes use the same sub-domains. The shape functions based on the MLS approximation, in an integration domain, have a single type of a rational function, which reduces the difficulty of numerical integration to evaluate the weak form. The present method is very useful in an adaptive calculation, because the secondary nodes can be easily added and/or moved without an additional mesh. The essential boundary conditions can be imposed exactly, and non-convex boundaries can be treated without special techniques. Several numerical examples are presented to illustrate the performance of the present method.

Atluri, Satya N.↗

NASA NTRS2011

Feature Acquisition with Imbalanced Training Data

This work considers cost-sensitive feature acquisition that attempts to classify a candidate datapoint from incomplete information. In this task, an agent acquires features of the datapoint using one or more costly diagnostic tests, and eventually ascribes a classification label. A cost function describes both the penalties for feature acquisition, as well as misclassification errors. A common solution is a Cost Sensitive Decision Tree (CSDT), a branching sequence of tests with features acquired at interior decision points and class assignment at the leaves. CSDT's can incorporate a wide range of diagnostic tests and can reflect arbitrary cost structures. They are particularly useful for online applications due to their low computational overhead. In this innovation, CSDT's are applied to cost-sensitive feature acquisition where the goal is to recognize very rare or unique phenomena in real time. Example applications from this domain include four areas. In stream processing, one seeks unique events in a real time data stream that is too large to store. In fault protection, a system must adapt quickly to react to anticipated errors by triggering repair activities or follow- up diagnostics. With real-time sensor networks, one seeks to classify unique, new events as they occur. With observational sciences, a new generation of instrumentation seeks unique events through online analysis of large observational datasets. This work presents a solution based on transfer learning principles that permits principled CSDT learning while exploiting any prior knowledge of the designer to correct both between-class and withinclass imbalance. Training examples are adaptively reweighted based on a decomposition of the data attributes. The result is a new, nonparametric representation that matches the anticipated attribute distribution for the target events.

Thompson, David R.↗

NASA NTRSDate not supplied

High Fidelity Adaptively Refined CFD and Reduced Order Models of a High Aspect Ratio Aeroelastic Wing Wind-Tunnel Model

The NASA Advanced Air Transport Technology (AATT) goal of reduced fuel burn for transport aircraft has led to the NASA N+3 High Aspect Ratio Wing (HARW) subproject. This project requires identifying, developing, and demonstrating key technologies and integrated multidisciplinary solutions to enable a safe, high performance, aeroelastic wing. Since this aircraft will have a high aspect ratio wing, aeroelasticity is expected to be a major issue in the design. In this paper high fidelity computational fluid dynamics (CFD) is performed with flow adapted meshes. A system identification of the aerodynamics is developed using both a multi-modal multi-sine time-marching and a multi-mode linear frequency domain method. GLA, MLA and flutter suppression simulations will be performed.

Robert Bartels↗

NASA NTRS2000

A Very High Order, Adaptable MESA Implementation for Aeroacoustic Computations

Since computational efficiency and wave resolution scale with accuracy, the ideal would be infinitely high accuracy for problems with widely varying wavelength scales. Currently, many of the computational aeroacoustics methods are limited to 4th order accurate Runge-Kutta methods in time which limits their resolution and efficiency. However, a new procedure for implementing the Modified Expansion Solution Approximation (MESA) schemes, based upon Hermitian divided differences, is presented which extends the effective accuracy of the MESA schemes to 57th order in space and time when using 128 bit floating point precision. This new approach has the advantages of reducing round-off error, being easy to program. and is more computationally efficient when compared to previous approaches. Its accuracy is limited only by the floating point hardware. The advantages of this new approach are demonstrated by solving the linearized Euler equations in an open bi-periodic domain. A 500th order MESA scheme can now be created in seconds, making these schemes ideally suited for the next generation of high performance 256-bit (double quadruple) or higher precision computers. This ease of creation makes it possible to adapt the algorithm to the mesh in time instead of its converse: this is ideal for resolving varying wavelength scales which occur in noise generation simulations. And finally, the sources of round-off error which effect the very high order methods are examined and remedies provided that effectively increase the accuracy of the MESA schemes while using current computer technology.

Dydson, Roger W.↗

NASA NTRS1993

A solution-adaptive hybrid-grid method for the unsteady analysis of turbomachinery

A solution-adaptive method for the time-accurate analysis of two-dimensional flows in turbomachinery is described. The method employs a hybrid structured-unstructured zonal grid topology in conjunction with appropriate modeling equations and solution techniques in each zone. The viscous flow region in the immediate vicinity of the airfoils is resolved on structured O-type grids while the rest of the domain is discretized using an unstructured mesh of triangular cells. Implicit, third-order accurate, upwind solutions of the Navier-Stokes equations are obtained in the inner regions. In the outer regions, the Euler equations are solved using an explicit upwind scheme that incorporates a second-order reconstruction procedure. An efficient and robust grid adaptation strategy, including both grid refinement and coarsening capabilities, is developed for the unstructured grid regions. Grid adaptation is also employed to facilitate information transfer at the interfaces between unstructured grids in relative motion. Results for grid adaptation to various features pertinent to turbomachinery flows are presented. Good comparisons between the present results and experimental measurements and earlier structured-grid results are obtained.

Mathur, Sanjay R.↗

NASA NTRS2007

Adaptive Modeling of the International Space Station Electrical Power System

Software simulations provide NASA engineers the ability to experiment with spacecraft systems in a computer-imitated environment. Engineers currently develop software models that encapsulate spacecraft system behavior. These models can be inaccurate due to invalid assumptions, erroneous operation, or system evolution. Increasing accuracy requires manual calibration and domain-specific knowledge. This thesis presents a method for automatically learning system models without any assumptions regarding system behavior. Data stream mining techniques are applied to learn models for critical portions of the International Space Station (ISS) Electrical Power System (EPS). We also explore a knowledge fusion approach that uses traditional engineered EPS models to supplement the learned models. We observed that these engineered EPS models provide useful background knowledge to reduce predictive error spikes when confronted with making predictions in situations that are quite different from the training scenarios used when learning the model. Evaluations using ISS sensor data and existing EPS models demonstrate the success of the adaptive approach. Our experimental results show that adaptive modeling provides reductions in model error anywhere from 80% to 96% over these existing models. Final discussions include impending use of adaptive modeling technology for ISS mission operations and the need for adaptive modeling in future NASA lunar and Martian exploration.

Thomas, Justin Ray↗

DOE OSTI2023

A Graph Dynamical neural network approach for decoding dynamical states in ferroelectrics.

Ferroelectric materials such as BaTiO 3 show tremendous potential for emerging advances in memory devices, particular neuromorphic type devices. High density of memory can be obtained by stabilising polar domain walls at the nanoscale, regions of discontinuity between the well-defined polarization order parameter, but little is known about what controls their structure and dynamics in real nanoscale materials. Indeed, chiral polar domain walls have been observed in heterogeneous ferroelectrics, such as oxygen-deficient BaTiO 3 , but very little is known about how such polar-domains walls interact with defects. Indeed, a critical understanding of how dynamics of domain-walls depend on point-defects is crucial to create engineered ferroelectric memory devices. For this work, we perform large-scale simulations of nansocale domain-wall dynamics in pristine and defective BaTiO 3 using reactive force-field developed by us earlier (Phys. Chem. Chem. Phys., 2019, 21, 18240–18249), and capture their dynamical dependence on point defects using a graph dynamical neural-network approach, which we adapted to interrogate solids with well-defined order-parameters, and implemented using Pytorch based libraries. Our machine learning (ML) approach goes beyond the traditional post-processing methods to capture both spatial and temporal heterogeneities of large-scale molecular dynamics simulations of complex defective ferroelectric oxide materials. We crucially find that isolated oxygen vacancies introduce very localized spatial regions (~1–2 unit-cell in length) that show slow dipole relaxation due to formation of defect-dipoles, and that these defect-dipoles in turn slow the intrinsic dynamics of domain walls. Further, the roughness of domain walls, also influenced by vacancies, introduce dynamic heterogeneity along the domain-wall. As such we find a novel mechanism by which quenched disorder due to defects introduce dynamic heterogeneity thereby influencing response to external fields (particularly time varying fields) in a ferroelectric. Our study also emphasizes the need for creating digital twins of dynamical quantities to achieve autonomous in operando control of nanoscale switching.

42 ENGINEERING↗

DOE OSTI2020

Structure of the respiratory MBS complex reveals iron-sulfur cluster catalyzed sulfane sulfur reduction in ancient life

Abstract Modern day aerobic respiration in mitochondria involving complex I converts redox energy into chemical energy and likely evolved from a simple anaerobic system now represented by hydrogen gas-evolving hydrogenase (MBH) where protons are the terminal electron acceptor. Here we present the cryo-EM structure of an early ancestor in the evolution of complex I, the elemental sulfur (S 0 )-reducing reductase MBS. Three highly conserved protein loops linking cytoplasmic and membrane domains enable scalable energy conversion in all three complexes. MBS contains two proton pumps compared to one in MBH and likely conserves twice the energy. The structure also reveals evolutionary adaptations of MBH that enabled S 0 reduction by MBS catalyzed by a site-differentiated iron-sulfur cluster without participation of protons or amino acid residues. This is the simplest mechanism proposed for reduction of inorganic or organic disulfides. It is of fundamental significance in the iron and sulfur-rich volcanic environments of early earth and possibly the origin of life. MBS provides a new perspective on the evolution of modern-day respiratory complexes and of catalysis by biological iron-sulfur clusters.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

Applying the FAIR Principles to computational workflows

Recent trends within computational and data sciences show an increasing recognition and adoption of computational workflows as tools for productivity and reproducibility that also democratize access to platforms and processing know-how. As digital objects to be shared, discovered, and reused, computational workflows benefit from the FAIR principles, which stand for Findable, Accessible, Interoperable, and Reusable. The Workflows Community Initiative’s FAIR Workflows Working Group (WCI-FW), a global and open community of researchers and developers working with computational workflows across disciplines and domains, has systematically addressed the application of both FAIR data and software principles to computational workflows. We present recommendations with commentary that reflects our discussions and justifies our choices and adaptations. These are offered to workflow users and authors, workflow management system developers, and providers of workflow services as guidelines for adoption and fodder for discussion. The FAIR recommendations for workflows that we propose in this paper will maximize their value as research assets and facilitate their adoption by the wider community.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1997

Solution adaptive Cartesian grid methods for aerodynamic flows with complex geometries

Cartesian methods for computational fluid dynamics are considered to offer an accurate and robust approach for the simulation of aerodynamic flows around geometrically complex bodies. A part of the ongoing research in this domain is reviewed with the aim of providing insight into the fundamental challenges faced by the practitioners of this approach, and a guide to further research. The integration schemes used in Cartesian solvers are similar to those used in other approaches. The emphasis is on the geometric algorithms, surface modeling and boundary conditions required to design a successful Cartesian mesh scheme.

Aftosmis, Michael J.↗

DOE OSTI2022

Newton trust-region methods with primary variable switching for simulating high temperature multiphase porous media flow

Coupling multiphase flow with energy transport due to high temperature heat sources introduces significant new challenges since boiling and condensation processes can lead to dry-out conditions with subsequent re-wetting. The transition between two-phase and single-phase behavior can require changes to the primary dependent variables adding discontinuities as well as extending constitutive nonlinear relations to extreme physical conditions. Practical simulations of large-scale engineered domains lead to Jacobian systems with a very large number of unknowns that must be solved efficiently using iterative methods in parallel on high-performance computers. Performance assessment of potential nuclear repositories, carbon sequestration sites and geothermal reservoirs can require numerous Monte-Carlo simulations to explore uncertainty in material properties, boundary conditions, and failure scenarios. Due to the numerical challenges, standard NR iteration may not converge over the range of required simulations and require more sophisticated optimization method like trust-region. In this study, we use the open-source simulator PFLOTRAN for the important practical problem of the safety assessment of future nuclear waste repositories in the U.S. DOE geologic disposal safety assessment Framework. The simulator applies the PETSc parallel framework and a backward Euler, finite volume discretization. We demonstrate failure of the conventional NR method and the success of trust-region modifications to Newton’s method for a series of test problems of increasing complexity. Trust-region methods essentially modify the Newton step size and direction under some circumstances where the standard NR iteration can cause the solution to diverge or oscillate. Furthermore, we show how the Newton Trust-Region method can be adapted for Primary Variable Switching (PVS) when the multiphase state changes due to boiling or condensation. The simulations with high-temperature heat sources which led to extreme nonlinear processes with many state changes in the domain did not converge with NR, but they do complete successfully with the trust-region methods modified for PVS. This implementation effectively decreased weeks of simulation time needing manual adjustments to complete a simulation down to a day. Finally, we show the strong scalability of the methods on a single node and multiple nodes in an HPC cluster.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRSDate not supplied

Standards Roadmapping for Mission Assurance in Commercial Spaceflight

This paper explores the application of established roadmapping approaches in standards planning and roadmap development to support mission assurance objectives. Mission assurance requires proactive planning across multiple domains including technical, schedule, cost, organizational, and policy. The intersection of these domains happens at the enterprise level, and enterprise systems engineering processes can systematically guide mission assurance activities, including development of standards roadmaps. Standards planning and development is a critical contributor to an effective mission assurance strategy, although standards are often developed and adopted reactively. Technology and standards planning is a key Enterprise Process Management activity, as identified by the Systems Engineering Body of Knowledge, and technology and standards roadmaps are strategic tools used to enable that process. Technology roadmapping has proven effective in guiding enterprise architecture and concept design, and standards planning supports enterprise requirements definition and management. However, there are limited methodologies to guide such standards planning activities. The application of technology roadmapping principles to standards planning formalizes the process and establishes a repeatable framework for standards development and integration. The framework development described in this paper fills a methodological gap by adapting proven roadmapping techniques to standards planning. Through a comparison of roadmapping methodologies and a case study analysis, this approach outlines a systematic process for anticipating and planning standards needs. In the case of developing a standards integration roadmap for NASA’s Office of Safety and Mission Assurance (OSMA), technology roadmapping principles were adapted to develop a roadmap framework to allow NASA and OSMA to define an agency-wide standards integration plan for commercial and industry safety and mission assurance standards. The outcome highlights key adaptations required for roadmapping in a standards context and describes the process steps to formalize a standards planning activity. The roadmapping framework adapted for standards development and integration enables more strategic, forward-looking standards planning and allows for better alignment between evolving capabilities and mission assurance requirements. Beyond its application to mission assurance, this methodology can be transferred to other domains requiring standards development, providing a systematic approach that integrates systems engineering principles into policy strategy and decision-making processes.

Enterprise Systems Engineering↗

DOE OSTI2023

Cost-constrained adaptive simulations of transient spray combustion in a gas turbine combustor

Predictive high-fidelity simulations of turbulent spray combustion must capture the combined effects of complex chemistry, multiphase evaporating flow and spray-flame interactions to achieve physical accuracy. Finite-rate chemistry (FRC) combined with a realistic chemical mechanism is a combustion model well-suited for this purpose, but has a high computational cost due to the large number and stiffness of transported chemical species. In contrast, flamelet-based models achieve lower cost by transporting a small number of quantities of reduced stiffness, but assumptions regarding local flame topology, boundary conditions and inter-phase coupling limit their physical accuracy. Recently, the Pareto-efficient combustion (PEC) framework was developed to dynamically assign combustion models based on local cost and accuracy metrics in gas-phase reacting flows. In this work, we extend this PEC framework to spray combustion through the rigorous analysis of the multiphase coupling terms in the governing equations. The derivation shows that spray evaporation causes errors in the prediction of species mass fractions for flamelet-based models due to the sensitivity of the local thermo-chemical state to changes in composition caused by fuel vaporization across combustion regimes present in practical spray combustion devices. Sub-model assignment is formulated as a multiple-choice knapsack problem, where computational cost is directly controlled through the fraction of the domain assigned to the FRC sub-model. The extended PEC formulation is applied to the simulation of a realistic rich-quench-lean gas turbine combustor at steady-state conditions, as well as transient operation resulting in lean blow-out (LBO). Analysis of transient simulations during LBO demonstrates the extended PEC formulation’s capacity to dynamically adapt to changing conditions within the combustor. Transient combustor dynamics are shown to approach convergence with limited increases in computational cost, while retaining substantial computational cost reduction compared to monolithic FRC simulations. Through PEC simulations with increasing fractions of the domain assigned to FRC, monolithic flamelet simulations are shown to over-predict flame stability during LBO. Finally, the extended PEC formulation is thus shown to overcome deficiencies of monolithic models by controlling modeling error for multiphase combustion modeling.

33 ADVANCED PROPULSION SYSTEMS↗

DOE OSTI2024

Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models

Abstract We propose masked particle modeling (MPM) as a self-supervised method for learning generic, transferable, and reusable representations on unordered sets of inputs for use in high energy physics (HEP) scientific data. This work provides a novel scheme to perform masked modeling based pre-training to learn permutation invariant functions on sets. More generally, this work provides a step towards building large foundation models for HEP that can be generically pre-trained with self-supervised learning and later fine-tuned for a variety of down-stream tasks. In MPM, particles in a set are masked and the training objective is to recover their identity, as defined by a discretized token representation of a pre-trained vector quantized variational autoencoder. We study the efficacy of the method in samples of high energy jets at collider physics experiments, including studies on the impact of discretization, permutation invariance, and ordering. We also study the fine-tuning capability of the model, showing that it can be adapted to tasks such as supervised and weakly supervised jet classification, and that the model can transfer efficiently with small fine-tuning data sets to new classes and new data domains.

Heinrich, Lukas (ORCID:0000000240487584)↗

NASA NTRS1987

An adaptive finite element scheme for the Euler and Navier-Stokes equations

The implementation of two explicit finite element schemes for the solution of the compressible Euler and Navier-Stokes equations is presented. The schemes can be employed with general unstructured triangular meshes in two dimensions. Either scheme can therefore be used as the basic solver in a solution adaptive mesh procedure in a direct manner. The particular adaptive approach which is advocated here is intended for the solution of steady state problems only and involves an adaptive regeneration of the grid at prescribed stages during the false transient. The grid regeneration is accomplished by a mesh generator which has the capability of generating triangular grids over computational domains of arbitrary shape. The procedure is illustrated by solving transonic flows over multi-airfoil configurations and high speed flows, involving shock interactions, past circular cylinders.

Morgan, K.↗

NASA NTRS1987

A linear quadratic Gaussian with loop transfer recovery proximity operations autopilot for spacecraft

An automatic control scheme for spacecraft proximity operations is presented. The controller is capable of holding the vehicle at a prescribed location relative to a target, or maneuvering it to a different relative position using straight line-of-sight translations. The autopilot uses a feedforward loop to initiate and terminate maneuvers, and for operations at nonequilibrium set-points. A multivariate feedback loop facilitates precise position and velocity control in the presence of sensor noise. The feedback loop is formulated using the Linear Quadratic Gaussian (LQG) with Loop Transfer Recovery (LTR) design procedure. Linear models of spacecraft dynamics, adapted from Clohessey-Wiltshire Equations, are augmented and loop shaping techniques are applied to design a target feedback loop. The loop transfer recovery procedure is used to recover the frequency domain properties of the target feedback loop. The resulting compensator is integrated into an autopilot which is tested in a high fidelity Space Shuttle Simulator. The autopilot performance is evaluated for a variety of proximity operations tasks envisioned for future Shuttle flights.

Chen, George T.↗

NASA NTRS1992

Intelligent data management for real-time spacecraft monitoring

Real-time AI systems have begun to address the challenge of restructuring problem solving to meet real-time constraints by making key trade-offs that pursue less than optimal strategies with minimal impact on system goals. Several approaches for adapting to dynamic changes in system operating conditions are known. However, simultaneously adapting system decision criteria in a principled way has been difficult. Towards this end, a general technique for dynamically making such trade-offs using a combination of decision theory and domain knowledge has been developed. Multi-attribute utility theory (MAUT), a decision theoretic approach for making one-time decisions is discussed and dynamic trade-off evaluation is described as a knowledge-based extension of MAUT that is suitable for highly dynamic real-time environments, and provides an example of dynamic trade-off evaluation applied to a specific data management trade-off in a real-world spacecraft monitoring application.

Schwuttke, Ursula M.↗