Search NASA⌕ Search

SEARCH · Search NASA

Results for “Bayesian learning”

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.

841 records · Page 47

Inferring fracture dilation and shear slip from surface deformation utilising trained surrogate models

An important task in energy and CO 2 storage (sequestration) in the subsurface is to verify that the surrounding fractures and faults are not activated, acting as leakage pathways. This is achievable through effective and efficient Measurement, Monitoring and Verification (MMV) plans. In this work, two surrogate models are trained to captures dilation (opening) and shear deformation of fractures, and the associated surface deformation. The trained surrogate model, based on conditional Generative-Adversarial Networks (cGAN) receives fracture apertures from dilational fractures together with fracture slips from shear fractures and predicts the combined surface deformation. An inversion algorithm based on Bayesian framework is proposed to identify the geometry of both types of fractures, as well as volume of dilational fractures and deformation moment induced by shear fractures, all from the measured surface deformation data. The inversion algorithm utilises the Differential Evolution (DE) optimisation technique that has the superior performance in finding the global minimum of cost function. The proposed surrogate-assisted inversion successfully inferred the unknown dip, dip direction and the volume of the dilational fractures as well as the induced deformation moment in shear fractures. The model was further tested for the inversion of a field hydraulic fracturing tilt dataset applying different scenarios with varying unknowns to show the model's performance, as well as incorporating shear deformation for better match with the observed data.

Dilation and shear↗

Domain Knowledge Guided Bayesian Optimization For Autonomous Alignment Of Complex Scientific Instruments

Bayesian Optimization (BO) is a powerful tool for optimizing complex non-linear systems. However, its performance degrades in high-dimensional problems with tightly coupled parameters and highly asymmetric objective landscapes, where rewards are sparse. In such needle-in-a-haystack scenarios, even advanced methods like trust-region BO (TurBO) often lead to unsatisfactory results. We propose a domain knowledge guided Bayesian Optimization approach, which leverages physical insight to fundamentally simplify the search problem by transforming coordinates to decouple input features and align the active subspaces with the primary search axes. We demonstrate this approach's efficacy on a challenging 12-dimensional, 6-crystal Split-and-Delay optical system, where conventional approaches, including standard BO, TuRBO and multi-objective BO, consistently led to unsatisfactory results. When combined with an reverse annealing exploration strategy, this approach reliably converges to the global optimum. The coordinate transformation itself is the key to this success, significantly accelerating the search by aligning input co-ordinate axes with the problem's active subspaces. As increasingly complex scientific instruments, from large telescopes to new spectrometers at X-ray Free Electron Lasers are deployed, the demand for robust high-dimensional optimization grows. Our results demonstrate a generalizable paradigm: leveraging physical insight to transform high-dimensional, coupled optimization problems into simpler representations can enable rapid and robust automated tuning for consistent high performance while still retaining current optimization algorithms.

FOS: Computer and information sciences↗

Genetic Algorithm for Optimization of Neural Networks for Bayesian Inference of Model Uncertainty

The objective of this work was to develop a genetic optimization algorithm that can design a neural network capable of producing uncertainty estimates along with predictions. This algorithm is necessary because the inclusion of uncertainty modeling in a neural network greatly complicates the network’s design space, making the development of a converging model extremely difficult and time consuming. The genetic algorithm presented in this work uses a number of value ranges for various configurable neural network parameters to create a randomly generated population of network architectures. The initially generated population is then evolved over the course of several generations, with the best performing models breeding to produce novel network configurations. Mutations are randomly applied to the network designs to facilitate the development of adaptations beneficial to the task being performed. An experiment was conducted to validate the proposed algorithm, in which the genetic optimizer was tasked with producing a neural network capable of predicting the sound pressure level (SPL) resulting from jet-surface interaction (JSI) noise. The data used for this task was generated at the NASA Glenn Research Center in the Aero-Acoustic Propulsion Laboratory. Starting with an initial population size of 35 randomly generated networks, and evolved over the course of 10 generations, the genetic algorithm produced a design able to predict SPL as a result of JSI noise within 0.272 dB, on average.

Genetic algorithm↗

Multidisciplinary Design Optimization and Analysis of an Open Rotor Stage: Part 1

Successful design of open rotor propulsors requires effective analysis across multiple disciplines, including aerodynamics, acoustics, and structures. A viable design must not only be efficient but must also produce an acceptable level of noise and meet all static and dynamic structural requirements. For design and optimization, this is especially challenging because running high fidelity analyses is resource-intensive, and optimizing a design may require many thousands of cases to be analyzed. For this reason, the NASA team has applied design methodology that utilizes low-cost aerodynamic methods, machine learning models, and high-fidelity analyses when necessary. This includes standard two-dimensional methods such as throughflow analysis and quasi-3D blade-to-blade CFD, as well as some newly developed methods. Optimization using 3D CFD is necessary to maximize performance, and this is considered as well. All optimizations are carried out subject to structural constraints evaluated using finite element analysis. Doing this accurately requires a robust trunnion design, capable of pitching the blade between cruise and takeoff conditions while maintaining acceptable factor of safety. Hot to cold analysis must also be applied in order to correctly determine the as-manufactured shape. For acoustics, the unsteady pressures on the blade surfaces must be predicted, and this can be done either through full-annulus unsteady CFD or through a nonlinear harmonic method (NLH). NLH can run much faster, allowing some acoustic considerations to be made earlier in the design process. The design process is ongoing, and this presentation will review the current status and planned next steps. This part of the talk will focus on aerodynamic performance and be followed by a talk on structures and acoustics.

Design↗

Sequential Design of Experiments to Maximize Learning from Carbon Capture Pilot Plant Testing

Pilot plant test campaigns can be expensive and time-consuming. Therefore, it is of interest to maximize the amount of learning and the efficiency of the test campaign given the limited number of experiments that can be conducted. This work investigates the use of sequential design of experiments (SDOE) to overcome these challenges by demonstrating its usefulness for a recent solvent-based CO2 capture plant test campaign. Unlike traditional design of experiments methods, SDOE regularly uses information from ongoing experiments to determine the optimum locations in the design space for subsequent runs within the same experiment. However, there are challenges that need to be addressed, including reducing the high computational burden to efficiently update the model, and the need to incorporate the methodology into a computational tool. We address these challenges by applying SDOE in combination with a software tool, the Framework for Optimization, Quantification of Uncertainty and Surrogates (FOQUS) (Miller et al., 2014a, 2016, 2017). The results of applying SDOE on a pilot plant test campaign for CO2 capture suggests that relative to traditional design of experiments methods, SDOE can more effectively reduce the uncertainty of the model, thus decreasing technical risk. Future work includes integrating SDOE into FOQUS and using SDOE to support additional large-scale pilot plant test campaigns.

20 FOSSIL-FUELED POWER PLANTS↗

Modeling to Improve the Risk Reduction Process for Command File Errors

The Jet Propulsion Laboratory has learned that even innocuous errors in the spacecraft command process can have significantly detrimental effects on a space mission. Consequently, such Command File Errors (CFE), regardless of their effect on the spacecraft, are treated as significant events for which a root cause is identified and corrected. A CFE during space mission operations is often the symptom of imbalance or inadequacy within the system that encompasses the hardware and software used for command generation as well as the human experts and processes involved in this endeavor. As we move into an era of increased collaboration with other NASA centers and commercial partners, these systems become more and more complex. Consequently, the ability to thoroughly model and analyze CFEs formally in order to reduce the risk they pose is increasingly important. In this paper, we summarize the results of applying modeling techniques previously developed to the DAWN flight project. The original models were built with the input of subject matter experts from several flight projects. We have now customized these models to address specific questions for the DAWN flight project and formulating use cases to address their unique mission needs. The goal of this effort is to enhance the project's ability to meet commanding reliability requirements for operations and to assist them in managing their Command File Errors.

spacecraft↗

Nonlinear Ensemble Filtering with Diffusion Models: Application to the Surface Quasigeostrophic Dynamics

The intersection between classical data assimilation methods and novel machine learning techniques has attracted significant interest in recent years. Here, we explore another promising solution in which diffusion models are used to formulate a robust nonlinear ensemble filter for sequential data assimilation. Unlike standard machine learning methods, the proposed ensemble score filter (EnSF) is completely training free and can efficiently generate a set of analysis ensemble members. Here, in this study, we apply the EnSF to a surface quasigeostrophic model and compare its performance against the popular local ensemble transform Kalman filter (LETKF), which makes Gaussian assumptions in the analysis step. Numerical tests demonstrate that EnSF maintains stable performance in the absence of localization and for a variety of experimental settings. We find that while LETKF maintains optimal performance in the case of linear observations of the entire state and a perfect model, EnSF shows improvements over LETKF when nonlinear observations are assimilated and the system is subject to unexpected model errors. A spectral decomposition of the analysis results in this nonlinear observation regime shows that the largest improvements over LETKF occur at large scales (small wavenumbers), where LETKF lacks sufficient ensemble spread. Overall, this initial application of EnSF to a geophysical model of intermediate complexity motivates further development of the algorithm for more realistic problems.

Artificial intelligence↗

Continuous Time Group Discovery in Dynamic Graphs

With the rise in availability and importance of graphs and networks, it has become increasingly important to have good models to describe their behavior. While much work has focused on modeling static graphs, we focus on group discovery in dynamic graphs. We adapt a dynamic extension of Latent Dirichlet Allocation to this task and demonstrate good performance on two datasets. Modeling relational data has become increasingly important in recent years. Much work has focused on static graphs - that is fixed graphs at a single point in time. Here we focus on the problem of modeling dynamic (i.e. time-evolving) graphs. We propose a scalable Bayesian approach for community discovery in dynamic graphs. Our approach is based on extensions of Latent Dirichlet Allocation (LDA). LDA is a latent variable model for topic modeling in text corpora. It was extended to deal with topic changes in discrete time and later in continuous time. These models were referred to as the discrete Dynamic Topic Model (dDTM) and the continuous Dynamic Topic Model (cDTM), respectively. When adapting these models to graphs, we take our inspiration from LDA-G and SSN-LDA, applications of LDA to static graphs that have been shown to effectively factor out community structure to explain link patterns in graphs. In this paper, we demonstrate how to adapt and apply the cDTM to the task of finding communities in dynamic networks. We use link prediction to measure the quality of the discovered community structure and apply it to two different relational datasets - DBLP author-keyword and CAIDA autonomous systems relationships. We also discuss a parallel implementation of this approach using Hadoop. In Section 2, we review LDA and LDA-G. In Section 3, we review the cDTM and introduce cDTMG, its adaptation to modeling dynamic graphs. We discuss inference for the cDTM-G and details of our parallel implementation in Section 4 and present its performance on two datasets in Section 5 before concluding in Section 6.

97 MATHEMATICS AND COMPUTING↗

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↗

Assessment of lumen degradation and remaining useful life of LEDs using particle filter

With the development of light-emitting diode (LED) technology, light emitting diodes system is becoming a popular light source in daily life and industry area. It has shown that Led from same factory and work under same working condition, may have significantly different behavior. Therefore, it is very important to learn the fail mechanisms, especially in the case of safety critical and harsh environment application. This paper focus on a prognostic health management (PHM) method based on the measurement of forward voltage and forward current of bare LED under harsh environment. In this paper, experiment has been done with ten samples. Ten pristine bare LEDs have been tested at 85°C while simultaneously being subjected to 85% humid environment. Pulse width modulation (PWM) control method has been employed to drive the bare LED in order to reduce the heat effect caused by forward current and high frequency (300HZ) data acquisition has been used to measure the peak forward voltage and forward current. Test to failure (lumen drops to 70 percent) data has been measured to study the effects of high temperature and humid environment loadings on the bare LED. Also, solid state cooling method with peltier cooler has been used to control the temperature of LED in the integrating sphere when take the measurement of lumen flux. The shift of forward voltage forward current curve and lumen degradation has been recorded to help build the fail model and predicted the remaining useful life. In this method, particle filter has been employed to predict the remaining useful life (RUL) of the bare LED and give us a whole picture how Led system fails. Result shows that predication of remaining useful life of Led, made by the particle filter model works under reasonable limit, and hence this method can be employed to predict the failure of Led caused by thermal and humid stress under harsh environment.

42 ENGINEERING↗

Residual-based error correction for neural operator accelerated infinite-dimensional Bayesian inverse problems

We explore using neural operators, or neural network representations of nonlinear maps between function spaces, to accelerate infinite-dimensional Bayesian inverse problems (BIPs) with models governed by nonlinear parametric partial differential equations (PDEs). Neural operators have gained significant attention in recent years for their ability to approximate the parameter-to-solution maps defined by PDEs using as training data solutions of PDEs at a limited number of parameter samples. The computational cost of BIPs can be drastically reduced if the large number of PDE solves required for posterior characterization are replaced with evaluations of trained neural operators. However, reducing error in the resulting BIP solutions via reducing the approximation error of the neural operators in training can be challenging and unreliable. We provide an a priori error bound result that implies certain BIPs can be ill-conditioned to the approximation error of neural operators, thus leading to inaccessible accuracy requirements in training. To reliably deploy neural operators in BIPs, we consider a strategy for enhancing the performance of neural operators: correcting the prediction of a trained neural operator by solving a linear variational problem based on the PDE residual. We show that a trained neural operator with error correction can achieve a quadratic reduction of its approximation error, all while retaining substantial computational speedups of posterior sampling when models are governed by highly nonlinear PDEs. The strategy is applied to two numerical examples of BIPs based on a nonlinear reaction–diffusion problem and deformation of hyperelastic materials. We demonstrate that posterior representations of the two BIPs produced using trained neural operators are greatly and consistently enhanced by error correction.

97 MATHEMATICS AND COMPUTING↗

Forest aboveground biomass estimation through integration of sentinel-2 and PALSAR-2 time series: assessing models trained on GEDI and field inventory benchmarks

Accurate and spatially explicit forest Aboveground Biomass (AGB) mapping through remote sensing is critical for quantifying terrestrial carbon stocks and informing effective forest management strategies. However, AGB estimation in dense forests with complex terrain remains challenging due to satellite sensor signal saturation problem (saturation issue occurs in high biomass forests), structural complexity, and limited ground truth for calibration. This study presents a novel framework that integrates multi-temporal Sentinel-2 optical imagery, ALOS PALSAR-2 Synthetic Aperture Radar (SAR) data, and topographic variables with explainable Machine Learning to map AGB across mountainous forests within subtropical and temperate oceanic climate zones of Mexico. We evaluate the effects of temporal granularity and sensor synergy by comparing multiple temporal inputs and sensor configurations (Sentinel-2, PALSAR-2, and their fusion), and assess model performance using two reference datasets: NASA GEDI LiDAR-derived biomass and Mexico’s National Forest and Soil Inventory (INFyS). Our results showed that models trained on INFyS consistently outperformed those trained on GEDI, highlighting limitations in GEDI’s reliability in biomass estimates within this study region. Furthermore, the integration of Sentinel-2 and PALSAR-2 provided improved predictions compared to single-sensor models, particularly when combined with temporally explicit yearly statistics. The best-performing model, which was trained on INFyS data, and considered both Sentinel-2 and PALSAR-2 yearly statistics, as well as topographic variables, achieved an R2 of 0.64, RMSE of 51.10 Mg/ha, and relative RMSE (rRMSE) of 58.69%. Explainable ML analysis identified Sentinel-2 spectral indices and topographic features as key predictors, while PALSAR-2 metrics provided complementary information, partially mitigating saturation effects in high-biomass areas. Specifically, integrating both sensors substantially improved AGB estimation in high biomass forest (≥200 Mg/ha), yielding 98% gains over optical-only model, with resulting estimates exceeding GEDI L4B by 29% and ESA-CCI-BIOMASS by 174%. Terrain-stratified analysis indicated close agreement with GEDI in low-slope areas, with increasing divergence as slope steepness increased, while estimates remained consistently higher than ESA-CCI-BIOMASS across all slope classes. The proposed approach advances multi-sensor fusion and temporal feature engineering for AGB mapping using open-access satellite datasets, providing a scalable and reproducible framework for annual biomass monitoring in topographically complex mountainous forests. The resulting 25 m resolution biomass product has the potential to provide spatially detailed information for forest monitoring and may support applications in carbon accounting and forest management.

54 ENVIRONMENTAL SCIENCES↗

Precision calibration of calorimeter signals in the ATLAS experiment using an uncertainty-aware neural network

The ATLAS experiment at the Large Hadron Collider explores the use of modern neural networks for a multi-dimensional calibration of its calorimeter signal defined by clusters of topologically connected cells (topo-clusters). The Bayesian neural network (BNN) approach not only yields a continuous and smooth calibration function that improves performance relative to the standard calibration but also provides uncertainties on the calibrated energies for each topo-cluster. The results obtained by using a trained BNN are compared to the standard local hadronic calibration and to a calibration provided by training a deep neural network. The uncertainties predicted by the BNN are interpreted in the context of a fractional contribution to the systematic uncertainties of the trained calibration. They are also compared to uncertainty predictions obtained from an alternative estimator employing repulsive ensembles.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗