Search NASA⌕ Search

SEARCH · Search NASA

Results for “operator inference”

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 595 records · Page 33

FPGA-accelerated SpeckleNN with SNL for real-time X-ray single-particle imaging

We present the implementation of a specialized version of our previously published unified embedding model, SpeckleNN, for real-time speckle pattern classification in X-ray Single-Particle Imaging (SPI), using the SLAC Neural Network Library (SNL) on an FPGA platform. This hardware realization transitions SpeckleNN from a prototypic model into a practical edge solution, optimized for running inference near the detector in high-throughput X-ray free-electron laser (XFEL) facilities, such as those found at the Linac Coherent Light Source (LCLS). To address the resource constraints inherent in FPGAs, we developed a more specialized version of SpeckleNN. The original model, which was designed for broader classification across multiple biological samples, comprised ~5.6 million parameters. The new implementation, while reducing the parameter count to 64.6K (a 98.8% reduction), focuses on maintaining the model's essential functionality for real-time operation, achieving an accuracy of 90%. Furthermore, we compressed the latent space from 128 to 50 dimensions. This implementation was demonstrated on the KCU1500 FPGA board, utilizing 71% of available DSPs, 75% of LUTs, and 48% of FFs, with an average power consumption of 9.4W according to the Vivado post-implementation report. The FPGA performed inference on a single image with a latency of 45.015 microseconds at a 200 MHz clock rate. In comparison, running the same inference on an NVIDIA A100 GPU resulted in an average power consumption of ~73W and an image processing latency of around 400 microseconds. Our FPGA-accelerated version of SpeckleNN demonstrated significant improvements, achieving an 8.9 × speedup and a 7.8 × reduction in power consumption compared to the GPU implementation. Key advancements include model specialization and dynamic weight loading through SNL, which eliminates the need for time-consuming FPGA design re-synthesis, allowing fast and continuous deployment of models (re)trained online. These innovations enable real-time adaptive classification and efficient vetoing of speckle patterns, making SpeckleNN more suited for deployment in XFEL facilities. This implementation has the potential to significantly accelerate SPI experiments and enhance adaptability to evolving experimental conditions.

47 OTHER INSTRUMENTATION↗

Beyond Point Estimates: Benchmarking Uncertainty Quantification Methods on the AION-1 Astronomical Foundation Model

Foundation models for astronomical surveys offer powerful learned representations that can be transferred to downstream regression tasks such as galaxy property estimation. However, point predictions alone are insufficient for scientific inference; reliable uncertainty quantification (UQ) is essential. We compare seven UQ methods on galaxy property regression using frozen AION-1 foundation-model embeddings, predicting redshift, stellar mass, stellar-population age, gas-phase metallicity, and specific star-formation rate, from Legacy Survey photometry/imaging and DESI spectra, with PROVABGS-derived labels. Distribution-free conformal methods achieve marginal coverage within $\sim$1 pp of the nominal 90% across all properties, while non-conformal baselines (Deep Ensembles, MC~Dropout) fail to calibrate reliably. Among conformal approaches, Conformalized Quantile Regression (CQR) delivers the best coverage in the bin with the poorest model predictions. More importantly, only the Locally Valid and Discriminative (LVD) framework -- particularly when operating on AION-1 embeddings -- also provides finite-sample \emph{local validity}, producing intervals that adapt to each galaxy's local prediction difficulty rather than relying on marginal guarantees alone. These results establish conformal prediction, and LVD in particular, as the preferred UQ framework for uncertainty-aware inference on foundation-model embeddings in astrophysics.

Tame-Narvaez, Karla [Fermilab] (ORCID:000000022249↗

Integration of facility modeling capabilities for nuclear nonproliferation analysis

Developing automated methods for data collection and analysis that can facilitate nuclear nonproliferation assessment is an important research area with significant consequences for the effective global deployment of nuclear energy. Facility modeling that can integrate and interpret observations collected from monitored facilities in order to ascertain their functional details will be a critical element of these methods. Although improvements are continually sought, existing facility modeling tools can characterize all aspects of reactor operations and the majority of nuclear fuel cycle processing steps, and include algorithms for data processing and interpretation. Assessing nonproliferation status is challenging because observations can come from many sources, including local and remote sensors that monitor facility operations, as well as open sources that provide specific business information about the monitored facilities, and can be of many different types. Although many current facility models are capable of analyzing large amounts of information, they have not been integrated in an analyst-friendly manner. This paper addresses some of these facility modeling capabilities and illustrates how they could be integrated and utilized for nonproliferation analysis. The inverse problem of inferring facility conditions based on collected observations is described, along with a proposed architecture and computer framework for utilizing facility modeling tools. After considering a representative sampling of key facility modeling capabilities, the proposed integration framework is illustrated with several examples.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Graphical models for optimal power flow

Optimal power flow (OPF) is the central optimization problem in electric power grids. Although solved routinely in the course of power grid operations, it is known to be strongly NP-hard in general, and weakly NP-hard over tree networks. In this paper, we formulate the optimal power flow problem over tree networks as an inference problem over a tree-structured graphical model where the nodal variables are low-dimensional vectors. We adapt the standard dynamic programming algorithm for inference over a tree-structured graphical model to the OPF problem. Combining this with an interval discretization of the nodal variables, we develop an approximation algorithm for the OPF problem. Further, we use techniques from constraint programming (CP) to perform interval computations and adaptive bound propagation to obtain practically efficient algorithms. Compared to previous algorithms that solve OPF with optimality guarantees using convex relaxations, our approach is able to work for arbitrary tree-structured distribution networks and handle mixed-integer optimization problems. Further, it can be implemented in a distributed message-passing fashion that is scalable and is suitable for “smart grid” applications like control of distributed energy resources. In conclusion, numerical evaluations on several benchmark networks show that practical OPF problems can be solved effectively using this approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Information and Statistics in Nuclear Experiment and Theory (ISNET)

As with all empirical sciences, nuclear physics operates in the virtuous cycle of the scientific method: observations inspire theoretical models; models lead to new predictions; predictions are tested in experiments; experiments lead to new observations; and so on. Evaluating what we are inferring, and how certain we are of it, is key to this process. These requirements, and a general interest in applying novel statistical, mathematical, and computational techniques, led to the formation of a dedicated research community entitled “Information and Statistics in Nuclear Experiment and Theory (ISNET)” (https://isnet-series.github.io/), which now includes more than 300 members. While the community’s interests lean toward nuclear theory, the unifying theme for this group is the inference of knowledge from data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Neural network accelerator for quantum control

Efficient quantum control is necessary for practical quantum computing implementations with current technologies. Conventional algorithms for determining optimal control parameters are computationally expensive, largely excluding them from use outside of the simulation. Existing hardware solutions structured as lookup tables are imprecise and costly. By designing a machine learning model to approximate the results of traditional tools, a more efficient method can be produced. Such a model can then be synthesized into a hardware accelerator for use in quantum systems. In this study, we demonstrate a machine learning algorithm for predicting optimal pulse parameters. This algorithm is lightweight enough to fit on a low-resource FPGA and perform inference with a latency of 175 ns and pipeline interval of 5 ns with > 0.99 gate fidelity. In the long term, such an accelerator could be used near quantum computing hardware where traditional computers cannot operate, enabling quantum control at a reasonable cost at low latencies without incurring large data bandwidths outside of the cryogenic environment.

43 PARTICLE ACCELERATORS↗

Absorber Column CFD model validation against PNNL’s device-scale absorber column on the LCFS unit

Absorber column has been widely used for CO 2 capture in the coal-fired power plants. High-fidelity CFD models play an important role in absorber column design and solvents optimization, which help enhance the CO 2 capture efficiency and reduce the operation cost. This report provides a comprehensive description of the development of CFD absorber models at two different levels, the design and implementation of PNNL’s device-scale absorber column experiment, and the methodology to combine the CFD results, experiment data, and Aspen model for a better understanding of the interface area in packed column. A composite model is firstly proposed in the Discrete Element Method (DEM) packing process, which can model the complex geometry of various packing elements. This generates a realistic packing pattern and accurate packing porosity ε and specific area a p compared to the actual values for absorber column used in LCFS. Two level of CFD absorber models were developed, namely the full-size column model (FCM) to simulate the entire packed column with a focus on the wall/entrance effects, and the representative column model (RCM) to simulate a section of column with a focus on the sensitivity study of interface area. A Design of Experiment (DoE) plan was developed to guide the collection of 100 run CFD data and 12 experiment runs. The 100 CFD runs were carried out in the RCM with Pro-Pak packing and cover a wide operation range and solvent properties. Impact of influential factors on the interface area in packed column were investigated in details. The CFD interface area was then combined with the experimental data and Aspen model to infer some information of the effective contact angle in the column. The accuracy and uncertainties in the interface area and contact angle can be quantified.

01 COAL, LIGNITE, AND PEAT↗

Building distributed rule-based systems using the AI Bus

The AI Bus software architecture was designed to support the construction of large-scale, production-quality applications in areas of high technology flux, running heterogeneous distributed environments, utilizing a mix of knowledge-based and conventional components. These goals led to its current development as a layered, object-oriented library for cooperative systems. This paper describes the concepts and design of the AI Bus and its implementation status as a library of reusable and customizable objects, structured by layers from operating system interfaces up to high-level knowledge-based agents. Each agent is a semi-autonomous process with specialized expertise, and consists of a number of knowledge sources (a knowledge base and inference engine). Inter-agent communication mechanisms are based on blackboards and Actors-style acquaintances. As a conservative first implementation, we used C++ on top of Unix, and wrapped an embedded Clips with methods for the knowledge source class. This involved designing standard protocols for communication and functions which use these protocols in rules. Embedding several CLIPS objects within a single process was an unexpected problem because of global variables, whose solution involved constructing and recompiling a C++ version of CLIPS. We are currently working on a more radical approach to incorporating CLIPS, by separating out its pattern matcher, rule and fact representations and other components as true object oriented modules.

Schultz, Roger D.↗

A First-Principles Model Based on Saturation of the Electron Cyclotron Drift Instability for Electron Transport in Hydrodynamics Simulations of Hall Thruster Plasmas

We present a computational model based on the hypothesis that the higher-than-expected electron mobility across magnetic field lines observed in Hall thrusters is due to the growth and eventual saturation of the electron cyclotron drift instability. The key improvement of this model with respect to previous work is that it decouples the saturation of the instability from the correlation that exists between mobility and magnitude of the wave perturbations in linear theory, thereby enabling first-principles simulations. The model has been incorporated in the 2-D (r-z) multi-fluids code Hall2De and simulations of the H6 thruster at 300 V, 20 A are numerically stable and achieve a steady-state solution at a computational cost that is not significantly higher than previous Hall2De simulations. The computed location of the acceleration region is found to be within 10% of the length of the acceleration channel compared to that inferred by experiments. The simulations also capture well the plasma gradients along the channel centerline of this thruster. To further establish the validity of this model, we plan to conduct simulations of other thrusters and/or operating conditions.

Lopez Ortega, Alejandro↗

Long-term variations in abundance and distribution of sulfuric acid vapor in the Venus atmosphere inferred from Pioneer Venus and Magellan radio occultation studies

Radio occultation experiments have been used to study various properties of planetary atmospheres, including pressure and temperature profiles, and the abundance profiles of absorbing constituents in those planetary atmospheres. However, the reduction of amplitude data from such experiments to determine abundance profiles requires the application of the inverse Abel transform (IAT) and numerical differentiation of experimental data. These two operations preferentially amplify measurement errors above the true signal underlying the data. A new technique for processing radio occultation data has been developed that greatly reduces the errors in the derived absorptivity and abundance profiles. This technique has been applied to datasets acquired from Pioneer Venus Orbiter radio occultation studies and more recently to experiments conducted with the Magellan spacecraft. While primarily designed for radar studies of the Venus surface, the high radiated power (EIRP) from the Magellan spacecraft makes it an ideal transmitter for measuring the refractivity and absorptivity of the Venus atmosphere by such experiments. The longevity of the Pioneer Venus Orbiter has made it possible to study long-term changes in the abundance and distribution of sulfuric acid vapor, H2SO4(g), in the Venus atmosphere between 1979 and 1992. The abundance of H2SO4(g) can be inferred from vertical profiles of 13-cm absorptivity profiles retrieved from radio occultation experiments. Data from 1979 and 1986-87 suggest that the abundance of H2SO4(g) at latitudes northward of 70 deg decreased over this time period. This change may be due to a period of active volcanism in the late 1970s followed by a relative quiescent period, or some other dynamic process in the Venus atmosphere. While the cause is not certain, such changes must be incorporated into dynamic models of the Venus atmosphere. Potentially, the Magellan spacecraft will extend the results of Pioneer Venus Orbiter and allow the continued monitoring of the abundance of distribution of H2SO4(g) in the Venus atmosphere, as well as other interesting atmospheric properties. Without such measurements it will be difficult to address other issues such as the short-term spatial variability of the abundance of H2SO4(g) at similar latitudes in Venus atmosphere, and the identities of particles responsible for large-scale variations observed in NIR images.

Jenkins, J. M.↗

Deep Learning Based Superconducting Radio-Frequency Cavity Fault Classification at Jefferson Laboratory

This work investigates the efficacy of deep learning (DL) for classifying C100 superconducting radio-frequency (SRF) cavity faults in the Continuous Electron Beam Accelerator Facility (CEBAF) at Jefferson Lab. CEBAF is a large, high-power continuous wave recirculating linac that utilizes 418 SRF cavities to accelerate electrons up to 12 GeV. Recent upgrades to CEBAF include installation of 11 new cryomodules (88 cavities) equipped with a low-level RF system that records RF time-series data from each cavity at the onset of an RF failure. Typically, subject matter experts (SME) analyze this data to determine the fault type and identify the cavity of origin. This information is subsequently utilized to identify failure trends and to implement corrective measures on the offending cavity. Manual inspection of large-scale, time-series data, generated by frequent system failures is tedious and time consuming, and thereby motivates the use of machine learning (ML) to automate the task. This study extends work on a previously developed system based on traditional ML methods (Tennant and Carpenter and Powers and Shabalina Solopova and Vidyaratne and Iftekharuddin, Phys. Rev. Accel. Beams, 2020, 23, 114601), and investigates the effectiveness of deep learning approaches. The transition to a DL model is driven by the goal of developing a system with sufficiently fast inference that it could be used to predict a fault event and take actionable information before the onset (on the order of a few hundred milliseconds). Because features are learned, rather than explicitly computed, DL offers a potential advantage over traditional ML. Specifically, two seminal DL architecture types are explored: deep recurrent neural networks (RNN) and deep convolutional neural networks (CNN). We provide a detailed analysis on the performance of individual models using an RF waveform dataset built from past operational runs of CEBAF. In particular, the performance of RNN models incorporating long short-term memory (LSTM) are analyzed along with the CNN performance. Furthermore, comparing these DL models with a state-of-the-art fault ML model shows that DL architectures obtain similar performance for cavity identification, do not perform quite as well for fault classification, but provide an advantage in inference speed.

97 MATHEMATICS AND COMPUTING↗

Artificial Reasoning System for Symptom-Based Conditional Failure Probability Estimation Using Bayesian Network

Advances in nuclear power technologies require enhanced capabilities for operator advice and autonomous control. One of the first tasks in the development of such capabilities is the formulation of symptom-based conditional failure probabilities for structures, systems, and components (SSCs) of interest, for which the primary goal is to aid plant personnel in deducing the probabilistic performance status of the monitored SSCs and in detecting impending faults/failure. The task of conditional failure probability estimation is a bidirectional inference problem and shall be logically tackled by the Bayesian network (BN) approach. As a knowledge-based artificial intelligence tool and a probabilistic graphical model, BN offers the capability of reasoning under uncertainty and graphical representation emulating the physical behavior of the target SSC. This paper provides a systematic overview of the BN technique and the software tools for handling implementation of BN models, along with the associated knowledge representation and reasoning paradigm. Both operational data and expert judgement can be readily incorporated into the knowledge base of a BN model. The challenges with data availability are highlighted, and the general approach to target SSC identification is presented. Our focus is upon failure-prone and risk-important balance of plant assets, especially cases having strong operator involvement. An exemplary case study on the failure of a motor-driven centrifugal pump is also conducted to demonstrate the usefulness and technical feasibility of the proposed artificial reasoning system using an expert system shell.

Zhao, Xingang↗

Quantum Graphical Models and Belief Propagation

Belief Propagation algorithms acting on Graphical Models of classical probability distributions, such as Markov Networks, Factor Graphs and Bayesian Networks, are amongst the most powerful known methods for deriving probabilistic inferences amongst large numbers of random variables. This paper presents a generalization of these concepts and methods to the quantum case, based on the idea that quantum theory can be thought of as a noncommutative, operator-valued, generalization of classical probability theory. Some novel characterizations of quantum conditional independence are derived, and definitions of Quantum n-Bifactor Networks, Markov Networks, Factor Graphs and Bayesian Networks are proposed. The structure of Quantum Markov Networks is investigated and some partial characterization results are obtained, along the lines of the Hammersley-Clifford theorem. A Quantum Belief Propagation algorithm is presented and is shown to converge on 1-Bifactor Networks and Markov Networks when the underlying graph is a tree. The use of Quantum Belief Propagation as a heuristic algorithm in cases where it is not known to converge is discussed. Applications to decoding quantum error correcting codes and to the simulation of many-body quantum systems are described.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Tip-based source of femtosecond electron pulses at 30 keV

We present a nano-scale photoelectron source, optimized for ultrashort pulse durations and well-suited for time-resolved diffraction and advanced laser acceleration experiments. A tungsten tip of several-ten-nanometers diameter mounted in a suppressor-extractor electrode configuration allows the generation of 30 keV electron pulses with an estimated pulse duration of 9 fs (standard deviation; 21 fs full width at half maximum) at the gun exit. We infer the pulse duration from particle tracking simulations, which are in excellent agreement with experimental measurements of the electron-optical properties of the source in the spatial domain. We also demonstrate femtosecond-laser triggered operation of the apparatus. The temporal broadening of the pulse upon propagation to a diffraction sample can be greatly reduced by collimating the beam. Besides the short electron pulse duration, a tip-based source is expected to feature a large transverse coherence and a nanometric emittance.

07 ISOTOPE AND RADIATION SOURCES↗

Differentially Private Synthesis and Sharing of Network Data Via Bayesian Exponential Random Graph Models

Abstract Network data often contain sensitive relational information. One approach to protecting sensitive information while offering flexibility for network analysis is to share synthesized networks based on the information in originally observed networks. We employ differential privacy (DP) and exponential random graph models (ERGMs) and propose the DP-ERGM method to synthesize network data. We apply DP-ERGM to two real-world networks. We then compare the utility of synthesized networks generated by DP-ERGM, the DyadWise Randomized Response (DWRR) approach, and the Synthesis through Conditional distribution of Edge given nodal Attribute (SCEA) approach. In general, the results suggest that DP-ERGM preserves the original information significantly better than two other approaches in network structural statistics and inference for ERGMs and latent space models. Furthermore, DP-ERGM satisfies node DP through modeling the global network structure with ERGM, a stronger notion of privacy than the edge DP under which DWRR and SCEA operate.

graph synthesis↗

Bayesian data assimilation for stochastic multiscale models of transport in porous media.

We investigate Bayesian techniques that can be used to reconstruct field variables from partial observations. In particular, we target fields that exhibit spatial structures with a large spectrum of lengthscales. Contemporary methods typically describe the field on a grid and estimate structures which can be resolved by it. In contrast, we address the reconstruction of grid-resolved structures as well as estimation of statistical summaries of subgrid structures, which are smaller than the grid resolution. We perform this in two different ways (a) via a physical (phenomenological), parameterized subgrid model that summarizes the impact of the unresolved scales at the coarse level and (b) via multiscale finite elements, where specially designed prolongation and restriction operators establish the interscale link between the same problem defined on a coarse and fine mesh. The estimation problem is posed as a Bayesian inverse problem. Dimensionality reduction is performed by projecting the field to be inferred on a suitable orthogonal basis set, viz. the Karhunen-Loeve expansion of a multiGaussian. We first demonstrate our techniques on the reconstruction of a binary medium consisting of a matrix with embedded inclusions, which are too small to be grid-resolved. The reconstruction is performed using an adaptive Markov chain Monte Carlo method. We find that the posterior distributions of the inferred parameters are approximately Gaussian. We exploit this finding to reconstruct a permeability field with long, but narrow embedded fractures (which are too fine to be grid-resolved) using scalable ensemble Kalman filters; this also allows us to address larger grids. Ensemble Kalman filtering is then used to estimate the values of hydraulic conductivity and specific yield in a model of the High Plains Aquifer in Kansas. Strong conditioning of the spatial structure of the parameters and the non-linear aspects of the water table aquifer create difficulty for the ensemble Kalman filter. We conclude with a demonstration of the use of multiscale stochastic finite elements to reconstruct permeability fields. This method, though computationally intensive, is general and can be used for multiscale inference in cases where a subgrid model cannot be constructed.

54 ENVIRONMENTAL SCIENCES↗

Improving Limited Angle CT Reconstruction with a Robust GAN Prior

Limited angle CT reconstruction is an under-determined linear inverse problem that requires appropriate regularization techniques to be solved. In this work we study how pre-trained generative adversarial networks (GANs) can be used to clean noisy, highly artifact laden reconstructions from conventional techniques, by effectively projecting onto the inferred image manifold. In particular, we use a robust version of the popularly used GAN prior for inverse problems, based on a recent technique called corruption mimicking, that significantly improves the reconstruction quality. The proposed approach operates in the image space directly, as a result of which it does not need to be trained or require access to the measurement model, is scanner agnostic, and can work over a wide range of sensing scenarios.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

The Martian surface as imaged, sampled, and analyzed by the Viking landers

Data collected by two Viking landers are analyzed. Attention is given to the characteristics of the surface inferred from Lander imaging and meteorology data, physical and magnetic properties experiments, and both inorganic and organic analyses of Martian samples. Viking Lander 1 touched down on Chryse Planitia on July 20, 1976 and continued to operate for 2252 sols, until November 20, 1982. Lander 2 touched down about 6500 km away from Lander 1, on Utopia Planitia on September 3, 1976. The chemical compositions of sediments at the two landing sites are similar, suggesting an aeolian origin. The compositions suggest an iron-rich rock an are matched by various clays and salts.

Arvidson, Raymond E.↗