Search NASA⌕ Search

SEARCH · Search NASA

Results for “Process optimization and mapping”

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 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Benders Decomposition Using Graph Modeling and Multi-Parametric Programming

Benders decomposition is a widely used method for solving large and structured optimization problems, but its performance is affected by the repeated solution of subproblems. We propose a flexible and modular algorithmic framework for accelerating Benders decomposition. Specifically, we express the problem structure by using a graph-theoretic modeling abstraction in which nodes represent optimization subproblems and edges represent connectivity between subproblems. A key innovation of our approach is that we embed multiparametric programming (mp) surrogates for node subproblems, which maps the exact analytical map of the subproblem solution space. The use of mp surrogates allows us to replace subproblem solves with fast look-ups and function evaluations for primal and dual variables during the iterative Benders process. We formally show the equivalence between classical Benders cuts and those derived from the mp solution. We implement our framework in the open-source PlasmoBenders.jl software package. To demonstrate the capabilities of the proposed framework, we apply it to a two-stage stochastic programming problem, which aims to make optimal capacity expansion decisions under market uncertainty. We evaluate both single-cut and multicut variants of Benders decomposition and show that the use of mp surrogates achieves substantial speedups in subproblem solve time, while preserving the convergence guarantees of Benders decomposition. We highlight advantages in solution analysis and interpretability that is enabled by mp critical region tracking; specifically, we show that these reveal how decisions evolve geometrically across the Benders search. Our results aim to demonstrate that combining surrogate modeling with graph modeling offers a promising and extensible foundation for structure-exploiting decomposition. In addition, by decomposing the problem into more tractable subproblems, the proposed approach also aims to overcome scalability issues of mp. Finally, the use of mp surrogates provides a unifying and modular optimization framework that enables the representation of heterogeneous node subproblems as modeling objects with a homogeneous structure.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1999

The Lunar Prospector Mission: Final Results of Trajectory Design, Quasi-Frozen Orbits, Extended Mission Targeting, and Lunar Topography and Potential Models

The National Aeronautics and Space Administration (NASA) selected Lunar Prospector (LP) as one of the discovery missions to conduct solar system exploration science investigations. The mission was NASA's first lunar voyage to investigate key science objectives since Apollo and was launched in January 1998. In keeping with discovery program requirements to reduce total mission cost and utilize new technology, Lunar Prospector's mission design and control will focus on the use of innovative and proven trajectory analysis programs. As part of this effort, the Goddard Space Flight Center and the Ames Research Center became partners in the Lunar Prospector trajectory team to provide the trajectory analysis and orbit determination support. At the end of 1998, Lunar Prospector completed its one-year primary mission at 100-km altitude above the lunar surface. On December 19, 1998, Lunar Prospector entered its extended mission phase. The mission orbit was lowered from 100 km to a mean altitude of 40 km. Due to lunar potential effects, the altitude of Lunar Prospector varied from 25 to 55 km above the mean lunar geoid. After one month at 40 km, the lunar potential model was updated based upon the new tracking data at 40 km. On January 15, 1999, the altitude was lowered again to a mean altitude of 30 km. The spherical altitude varied between 15 km and 45 km above the mean lunar geoid while the topographical altitude varied between 10 km and 50 km. Various means were employed to get accurate lunar surface elevation including Clementine altimetry and LOS analysis. Based upon the best available terrain maps, Lunar Prospector reached actual altitudes of 8 km above lunar mountains in the southern polar region. This extended mission phase of six months will enable LP to obtain science data up to 3 orders of magnitude better than at the mission orbit. At the end of the operations mission, LP was targeted for impact at a chosen location that allowed optical observation of the lunar ejecta as LP ended its mission at 1.6 km/sec. This paper details the trajectory design and orbit determination planning and actual results of the Lunar Prospector nominal and extended mission including maneuver design, eccentricity vs. argument of perigee evolution, topographical altitude estimation, and lunar potential modeling. This paper provides understanding of the quasi-frozen orbit design of the LP mission, the optimization process of lunar orbit targets, the impacts that the selected lunar potential models play, and discusses the feasibility of meeting the mission goals. Observed evolution of the Keplerian orbit elements are compared to the theoretical predictions using the latest lunar potential model available which incorporates the Lunar Prospector Doppler data. Mapping orbit maintenance maneuver design along with results of the actual maneuvers to maintain the orbital requirements are also presented.

Folta, David↗
DOE OSTI2024

Non-averaged single-molecule tertiary structures reveal RNA self-folding through individual-particle cryo-electron tomography

Large-scale and continuous conformational changes in the RNA self-folding process present significant challenges for structural studies, often requiring trade-offs between resolution and observational scope. Here, we utilize individual-particle cryo-electron tomography (IPET) to examine the post-transcriptional self-folding process of designed RNA origami 6-helix bundle with a clasp helix (6HBC). By avoiding selection, classification, averaging, or chemical fixation and optimizing cryo-ET data acquisition parameters, we reconstruct 120 three-dimensional (3D) density maps from 120 individual particles at an electron dose of no more than 168 e - Å -2 , achieving averaged resolutions ranging from 23 to 35 Å, as estimated by Fourier shell correlation (FSC) at 0.5. Each map allows us to identify distinct RNA helices and determine a unique tertiary structure. Statistical analysis of these 120 structures confirms two reported conformations and reveals a range of kinetically trapped, intermediate, and highly compacted states, demonstrating a maturation folding landscape likely driven by helix-helix compaction interactions.

36 MATERIALS SCIENCE↗
NASA NTRS1987

A post-processing system for automated rectification and registration of spaceborne SAR imagery

An automated post-processing system has been developed that interfaces with the raw image output of the operational digital SAR correlator. This system is designed for optimal efficiency by using advanced signal processing hardware and an algorithm that requires no operator interaction, such as the determination of ground control points. The standard output is a geocoded image product (i.e. resampled to a specified map projection). The system is capable of producing multiframe mosaics for large-scale mapping by combining images in both the along-track direction and adjacent cross-track swaths from ascending and descending passes over the same target area. The output products have absolute location uncertainty of less than 50 m and relative distortion (scale factor and skew) of less than 0.1 per cent relative to local variations from the assumed geoid.

Curlander, John C.↗
DOE OSTI2025

Fast-RF-Shimming: Accelerate RF shimming in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) offers an elevated signal-to-noise ratio (SNR), enabling exceptionally high spatial resolution that benefits both clinical diagnostics and advanced research. However, the jump to higher fields introduces complications, particularly transmit radiofrequency (RF) field ($B^{+}_{1}$) inhomogeneities, manifesting as uneven flip angles and image intensity irregularities. These artifacts can degrade image quality and impede broader clinical adoption. Traditional RF shimming methods, such as Magnitude Least Squares (MLS) optimization, effectively mitigate $B^{+}_{1}$ inhomogeneity, but remain time-consuming. Recent machine learning approaches, including RF Shim Prediction by Iteratively Projected Ridge Regression and other deep learning architectures, suggest alternative pathways. Although these approaches show promise, challenges such as extensive training periods, limited network complexity, and practical data requirements persist. In this paper, we introduce a holistic learning-based framework called Fast-RF-Shimming, which achieves a 5000 ​× ​speed-up compared to the traditional MLS method. In the initial phase, we employ random-initialized Adaptive Moment Estimation (Adam) to derive the desired reference shimming weights from multi-channel $B^{+}_{1}$ fields. Next, we train a Residual Network (ResNet) to map $B^{+}_{1}$ fields directly to the ultimate RF shimming outputs, incorporating the confidence parameter into its loss function. Finally, we design Non-uniformity Field Detector (NFD), an optional post-processing step, to ensure the extreme non-uniform outcomes are identified. Comparative evaluations with standard MLS optimization underscore notable gains in both processing speed and predictive accuracy, which indicates that our technique shows a promising solution for addressing persistent inhomogeneity challenges.

Deep learning↗
NASA NTRS1994

Fuzzy self-learning control for magnetic servo system

It is known that an effective control system is the key condition for successful implementation of high-performance magnetic servo systems. Major issues to design such control systems are nonlinearity; unmodeled dynamics, such as secondary effects for copper resistance, stray fields, and saturation; and that disturbance rejection for the load effect reacts directly on the servo system without transmission elements. One typical approach to design control systems under these conditions is a special type of nonlinear feedback called gain scheduling. It accommodates linear regulators whose parameters are changed as a function of operating conditions in a preprogrammed way. In this paper, an on-line learning fuzzy control strategy is proposed. To inherit the wealth of linear control design, the relations between linear feedback and fuzzy logic controllers have been established. The exercise of engineering axioms of linear control design is thus transformed into tuning of appropriate fuzzy parameters. Furthermore, fuzzy logic control brings the domain of candidate control laws from linear into nonlinear, and brings new prospects into design of the local controllers. On the other hand, a self-learning scheme is utilized to automatically tune the fuzzy rule base. It is based on network learning infrastructure; statistical approximation to assign credit; animal learning method to update the reinforcement map with a fast learning rate; and temporal difference predictive scheme to optimize the control laws. Different from supervised and statistical unsupervised learning schemes, the proposed method learns on-line from past experience and information from the process and forms a rule base of an FLC system from randomly assigned initial control rules.

Tarn, J. H.↗
NASA NTRS2017

Mapping the Limitations of Breakthrough Analysis in Fixed-Bed Adsorption

The separation of gases through adsorption plays an important role in the chemical processing industry, where the separation step is often the costliest part of a chemical process and thus worthy of careful study and optimization. This work developed a number of new, archival aspects on the computer simulations used for the refinement and design of these gas adsorption processes: 1. Presented a new approach to fit the undetermined heat and mass transfer coefficients in the axially dispersed plug flow equation and associated balance equations 2. Examined and described the conditions where non-physical simulation results can arise 3. Presented an approach to determine the limits of the axial dispersion and LDF mass transfer terms above which non-physical simulation results occur.

Knox, James Clinton↗
NASA NTRS2025

Mission-Maps For Outbound Cislunar Transfer Trajectories

This study quantifies the robustness and sensitivity of an outbound cislunar trajectory for a lunar lander in the form of mission-maps, or topological maps that allows either a computer program or mission designer to intuitively optimize the placement of critical outbound correction burns from the derived sensitivity data. The non-linear multi-body dynamics are applied to generate an outbound cislunar reference profile used by a linear covariance analysis (LinCov) tool to compute the expected Δv and trajectory dispersions due to the initial state uncertainty, sensor errors, maneuver execution errors, and disturbance accelerations along the outbound cislunar profile. The rapid performance analysis capabilities of LinCov are complimented with parallel processing techniques to evaluate hundreds and thousands of different translational burn locations, placements, and targeting constraints to identify the combination that minimizes the total Δv usage (nominal plus 3σ Δv) and trajectory dispersions at lunar orbit insertion. This study utilizes a generalized reference targeting algorithm to quickly assess the integrated closed-loop GN&C system performance due to different targeting configurations and constraints. The resulting mission maps provide an intuitive insight to ascertain each trajectory correction maneuver’s (TCM) sensitivity to different burn times along an outbound cislunar trajectory and quickly identify desirable engineering tradeoffs when performing analysis on the number and placement of these burns that nominally zero. Multiple mission maps are generated for a variety of different performance parameters that allow engineers to visually identify optimal solutions for trajectory correction maneuver placements, the number of correction burns, and the targeting constraints for each burn.

GN&C↗
DOE OSTI2024

Multivariate Bayesian Optimization of CoO Nanoparticles for CO 2 Hydrogenation Catalysis

The hydrogenation of CO 2 holds promise for transforming the production of renewable fuels and chemicals. However, the challenge lies in developing robust and selective catalysts for this process. Transition metal oxide catalysts, particularly cobalt oxide, have shown potential for CO 2 hydrogenation, with performance heavily reliant on crystal phase and morphology. Achieving precise control over these catalyst attributes through colloidal nanoparticle synthesis could pave the way for catalyst and process advancement. Yet, navigating the complexities of colloidal nanoparticle syntheses, governed by numerous input variables, poses a significant challenge in systematically controlling resultant catalyst features. We present a multivariate Bayesian optimization, coupled with a data-driven classifier, to map the synthetic design space for colloidal CoO nanoparticles and simultaneously optimize them for multiple catalytically relevant features within a target crystalline phase. The optimized experimental conditions yielded small, phase-pure rock salt CoO nanoparticles of uniform size and shape. These optimized nanoparticles were then supported on SiO 2 and assessed for thermocatalytic CO 2 hydrogenation against larger, polydisperse CoO nanoparticles on SiO 2 and a conventionally prepared catalyst. The optimized CoO/SiO 2 catalyst consistently exhibited higher activity and CH 4 selectivity (ca. 98%) across various pretreatment reduction temperatures as compared to the other catalysts. This remarkable performance was attributed to particle stability and consistent H* surface coverage, even after undergoing the highest temperature reduction, achieving a more stable catalytic species that resists sintering and carbon occlusion.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Using Flory–Huggins-informed human-in-the-loop Bayesian optimization to map the phase diagram of polymer blends

Mapping the phase diagram of polymer blends is an essential step in controlling the structure–property relationship of polymer-based materials. However, traditional grid-based approaches are inefficient and rely on subjective judgements for terminating the experimental campaign. Artificial intelligence-guided experimentation offers a compelling alternative, especially when data-driven decision-making is interfaced with established polymer thermodynamics to improve efficiency and interpretability. Here, we introduce a physics-informed Bayesian optimization approach to guide the mapping of the phase diagram of a model blend containing poly(methyl methacrylate) and poly(styrene-ran-acrylonitrile). Physical information is derived from a Flory–Huggins representation of the spinodal curve, which is integrated into the Bayesian optimization process as a structured prior mean that acts as a soft constraint. Implemented as a human-in-the-loop workflow, the approach leverages optical imaging of film cloudiness with iterative Gaussian process surrogate modeling and a parameter selection decision policy to identify the composition-temperature conditions for sequential iterations. Convergence of kernel and Flory–Huggins-based hyperparameters provided a stopping criterion, ensuring an objective and interpretable termination of the experimental campaign. The framework recovered the known lower critical solution temperature (∼160 °C), while increasing material efficiency through targeted sampling. This work establishes a proof-of-concept for the application of Bayesian optimization workflows to study polymer blend miscibility.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Simulation driven adaptive sampling for neutron-diffraction based strain mapping of additively manufactured parts

Neutron diffraction based strain mapping is a useful technique for measuring residual strains in additively manufactured (AM) metal parts. The measurement is traditionally done by scanning the sample in a point-wise raster pattern to extract the strain at each position. Since the overall scan can span several hours, adaptive sampling approaches using Bayesian optimization based on Gaussian process (BO-GP) regression have been introduced—demonstrating that even with a fraction of the typically made measurements the dominant strain patterns in the sample can be reconstructed. However, the parameters of the BO-GP algorithm have to be carefully chosen for best performance, and the movement time between arbitrary points can offset the time savings from a reduced number of measurement locations. In this paper, we propose algorithms to refine the BO-GP based methods by using simulations of strain patterns in AM parts based on the materials and the process used to print them. We demonstrate that the simulated strain patterns can be used to help choose better parameters for the BO-GP based framework—leading to low reconstruction error for the final strain pattern. Furthermore, we show that the strain mapping experiment can be initialized with a sampling pattern learnt from the simulation data and ordered to reduce movement time, dramatically enabling reduction in the overall time required to run the baseline BO-GP method.

Gaussian process regression↗
DOE OSTI2021

Multichannel Analysis of Surface Waves Accelerated (MASWAccelerated): Software for efficient surface wave inversion using MPI and GPUs

Multichannel Analysis of Surface Waves (MASW) is a technique frequently used in geotechnical engineering and engineering geophysics to infer 1D layered models of seismic shear wave velocities in the top tens to hundreds of meters of the subsurface. We aim to accelerate MASW calculations by capitalizing on modern computer hardware available in the workstations of most engineers: multiple cores and graphics processing units (GPUs). We propose new parallel and GPU accelerated algorithms for computing 1D MASW inversion, and provide software implementations in C using Message Passing Interface (MPI) and CUDA. These algorithms take advantage of sparsity that arises in the problem, and the work balance between processes considers typical data trends. We compare our methods to an existing open source Matlab MASW tool. Our serial C implementation achieves a 2x speedup over the Matlab software, and we continue to see improvements by parallelizing the problem with MPI. Here we see nearly perfect strong and weak scaling for uniform data, and improve strong scaling for realistic data by repartitioning the problem to process mapping. By utilizing GPUs available on most modern workstations, we observe an additional 1.3x speedup over the serial C implementation on the first use of the method. We typically repeatedly evaluate theoretical dispersion curves as part of an optimization procedure, and on the GPU the kernel can be cached for faster reuse on later runs. We observe a 3.2x speedup on the cached GPU runs compared to the serial C runs. This work is the first open-source parallel or GPU-accelerated software tool for MASW imaging, and should enable geotechnical engineers to fully utilize all computer hardware at their disposal.

58 GEOSCIENCES↗
NASA NTRS1984

SAR image enhancement via post-correlation signal processing

Seventeen interpreters ranked sets of computer-generated radar imagery to assess the value of post-correlation processing on the interpretability of SAR (synthetic aperture radar) imagery. The post-correlation processing evaluated amounts to a nonlinear mapping of the signal exiting a digital correlator and allows full use of signal bandwidth for improving the spatial resolution or for noise reduction. The results indicate that it is reasonable to hypothesize an optimal SAR presentation format for specific applications even though this study was too limited to be specific.

Matthews, N. D.↗
NASA NTRS1996

LCTV Holographic Imaging

Astronauts are required to interface with complex systems that require sophisticated displays to communicate effectively. Lightweight, head-mounted real-time displays that present holographic images for comfortable viewing may be the ideal solution. We describe an implementation of a liquid crystal television (LCTV) as a spatial light modulator (SLM) for the display of holograms. The implementation required the solution of a complex set of problems. These include field calculations, determination of the LCTV-SLM complex transmittance characteristics and a precise knowledge of the signal mapping between the LCTV and frame grabbing board that controls it. Realizing the hologram is further complicated by the coupling that occurs between the phase and amplitude in the LCTV transmittance. A single drive signal (a gray level signal from a framegrabber) determines both amplitude and phase. Since they are not independently controllable (as is true in the ideal SLM) one must deal with the problem of optimizing (in some sense) the hologram based on this constraint. Solutions for the above problems have been found. An algorithm has been for field calculations that uses an efficient outer product formulation. Juday's MEDOF 7 (Minimum Euclidean Distance Optimal Filter) algorithm used for originally for filter calculations has been successfully adapted to handle metrics appropriate for holography. This has solved the problem of optimizing the hologram to the constraints imposed by coupling. Two laboratory methods have been developed for determining an accurate mapping of framegrabber pixels to LCTV pixels. A friendly software system has been developed that integrates the hologram calculation and realization process using a simple set of instructions. The computer code and all the laboratory measurement techniques determining SLM parameters have been proven with the production of a high quality test image.

Knopp, Jerome↗
NASA NTRSDate not supplied

Method Development for In-situ Detection of Latent Herpesvirus DNA from Saliva using Nanopore Sequencing

Research toward latent herpesvirus reactivation has been intensively addressed through Space Shuttle and International Space Station (ISS) investigations. This work has provided the understanding that persistent reactivation of herpesviruses from asymptomatic crew can be detected through viral shedding in saliva, urine, and blood. Occasionally, viral reactivation from the latency stage can pose a threat to crew health (clinical manifestation) before, during, and after flight missions. Furthermore, previous work detailing correlations with immunity indicate that monitoring viral reactivation could be implemented to assess potential immune dysfunction. While in-flight monitoring is desirable, there is no well-established procedure or method for real-time evaluations, and research to date has relied on postflight, ground-based analysis. The development of portable molecular technologies like the miniPCR™ (miniPCR Bio) thermal cycler and the MinION™ sequencer (Oxford Nanopore Technologies) have confirmed that real-time monitoring is possible in extreme and low resource environments. These devices, combined with simple sample preparation methods, have been used to demonstrate bacterial identification onboard the ISS, as well as rapid viral detection in remote locations on Earth. The work here builds upon previous molecular advancements onboard the ISS toward the development and validation of a spaceflight-compatible method for viral detection from crew samples. Several herpesviruses can be detected in saliva, which provides a non-invasive means to collect samples for monitoring. While the basis for this method lies in previous spaceflight investigations, key points for method optimization include DNA extraction from saliva, viral primer selection, and bioinformatic processes for data analysis. To increase viral yield, numerous DNA extraction methods have been evaluated and will be discussed in detail. For initial development and testing, varicella-zoster virus (VZV) is being targeted though open reading frame 51 and 63 (ORF51, ORF63), as the replication origin-binding protein is highly expressed during latency. Optimization of the thermal cycling parameters has resulted in the ability to test the entire process. The full method has been tested with both viral VZV DNA standards and saliva spiked with varying concentrations of VZV. Viral sequence data were mapped to the reference sequence using minimap2. Prior to mapping, DNA sequencing reads were filtered for length and quality, barcodes were removed, and alignment identity calculated. Following further assessments, statistics were compared across multiple sequencing experiments and are being used to determine the success of the protocols. Forward work will include the incorporation of herpes simplex virus 1 (HSV-1) and Epstein-Barr virus (EBV) primers as well as the validation of results to the terrestrial qPCR standard assay. Upon full validation of the developed method, saliva will be collected from 20 healthy subjects and spiked with viral DNA. These samples will be split and assayed with the MinION and standard qPCR assay.

Hang N Nguyen↗
DOE OSTI2025

Optimization of an aerostructural machining process using physics-guided Bayesian stability modelling

Existing algorithms for predicting milling chatter have not been widely adopted in industry since they require specialized instruments to measure the stability inputs. This study describes how the machining process for a meter-scale aluminum aerostructure was optimized using a physics-guided Bayesian stability model. The study was performed in collaboration with an industrial partner on production machines to evaluate the practicality of the proposed method under real-world conditions. For each cutting tool, the Bayesian approach automatically selected a small number of cutting tests, which were monitored using a microphone to observe the chatter frequency. The algorithm learned the system dynamics, cutting forces, and stability map from these test results. A novel algorithm for predicting tool bending stress was incorporated into the test selection algorithm to avoid tool breakage. On average, each set of optimized cutting parameters required less than six tests to identify and were 97% more productive than baseline parameters from the cutting tool manufacturer. The machining program was then further optimized using commercial feedrate scheduling software to remove cutting force spikes and reduce air cutting time. Five components were machined using the optimized process. These results demonstrate the potential for physics-guided Bayesian models to improve productivity in industrial settings.

Cornelius, Aaron [UT Knoxville]↗
NASA NTRS1992

New approaches in interferometric SAR data processing

It is well established that interferometric synthetic aperture radar (SAR) images can be inverted to perform surface elevation mapping. Among the factors critical to the mapping accuracy are registration of the interfering SAR images and phase unwrapping. A novel registration algorithm is presented that determines the registration parameters through optimization. A new figure of merit is proposed that evaluates the registration result during the optimization. The phase unwrapping problem is approached through a new method involving fringe line detection. The algorithms are tested with two SEASAT SAR images of terrain near Yellowstone National Park. These images were collected on Seasat orbits 1334 and 1420, which were very close together in space, i.e., less than 100 m. The resultant elevation map is compared with the USGS digital terrain elevation model.

Lin, Qian↗
DOE OSTI2022

Data-driven wind turbine wake modeling via probabilistic machine learning

Wind farm design primarily depends on the variability of the wind turbine wake flows to the atmospheric wind conditions and the interaction between wakes. Physics-based models that capture the wake flow field with high-fidelity are computationally very expensive to perform layout optimization of wind farms, and, thus, data-driven reduced-order models can represent an efficient alternative for simulating wind farms. In this work, we use real-world light detection and ranging (LiDAR) measurements of wind-turbine wakes to construct predictive surrogate models using machine learning. Specifically, we first demonstrate the use of deep autoencoders to find a low-dimensional latent space that gives a computationally tractable approximation of the wake LiDAR measurements. Then, we learn the mapping between the parameter space and the (latent space) wake flow fields using a deep neural network. Additionally, we also demonstrate the use of a probabilistic machine learning technique, namely, Gaussian process modeling, to learn the parameter-space-latent-space mapping in addition to the epistemic and aleatoric uncertainty in the data. Finally, to cope with training large datasets, we demonstrate the use of variational Gaussian process models that provide a tractable alternative to the conventional Gaussian process models for large datasets. Furthermore, we introduce the use of active learning to adaptively build and improve a conventional Gaussian process model predictive capability. Overall, we find that our approach provides accurate approximations of the wind-turbine wake flow field that can be queried at an orders-of-magnitude cheaper cost than those generated with high-fidelity physics-based simulations.

Deep neural networks↗