Search NASA⌕ Search

SEARCH · Search NASA

Results for “residual networks”

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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

LossLens: Diagnostics for Machine Learning Through Loss Landscape Visual Analytics

Modern machine learning often relies on optimizing a neural network's parameters using a loss function to learn complex features. Beyond training, examining the loss function with respect to a network's parameters (i.e., as a loss landscape) can reveal insights into the architecture and learning process. While the local structure of the loss landscape surrounding an individual solution can be characterized using a variety of approaches, the global structure of a loss landscape, which includes potentially many local minima corresponding to different solutions, remains far more difficult to conceptualize and visualize. To address this difficulty, we introduce LossLens, a visual analytics framework that explores loss landscapes at multiple scales. LossLens integrates metrics from global and local scales into a comprehensive visual representation, enhancing model diagnostics. Here we demonstrate LossLens through two case studies: visualizing how residual connections influence a ResNet-20, and visualizing how physical parameters influence a physics-informed neural network (PINN) solving a simple convection problem.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Generative $β$-hairpin design using a residue-based physicochemical property landscape

De novo peptide design is a new frontier that has broad application potential in the biological and biomedical fields. Most existing models for de novo peptide design are largely based on sequence homology that can be restricted based on evolutionarily derived protein sequences and lack the physicochemical context essential in protein folding. Generative machine learning for de novo peptide design is a promising way to synthesize theoretical data that are based on, but unique from, the observable universe. In this study, we created and tested a custom peptide generative adversarial network intended to design peptide sequences that can fold into the -hairpin secondary structure. This deep neural network model is designed to establish a preliminary foundation of the generative approach based on physicochemical and conformational properties of 20 canonical amino acids, for example, hydrophobicity and residue volume, using extant structure-specific sequence data from the PDB. The beta generative adversarial network model robustly distinguishes secondary structures of hairpin from α helix and intrinsically disordered peptides with an accuracy of up to 96% and generates artificial -hairpin peptide sequences with minimum sequence identities around 31% and 50% when compared against the current NCBI PDB and nonredundant databases, respectively. These results highlight the potential of generative models specifically anchored by physicochemical and conformational property features of amino acids to expand the sequence-to-structure landscape of proteins beyond evolutionary limits.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Identifying Sample Provenance From SEM/EDS Automated Particle Analysis via Few-Shot Learning Coupled With Similarity Graph Clustering

Automated particle analysis (APA) provides a vast amount of compositional data via energy-dispersive X-ray spectroscopy along with size and shape data via scanning electron microscopy for individual particles in a sample. In many instances, APA data are leveraged to support identification of the source of a sample based on the detection of particles of a specific composition. Often, the particles that provide context make up a minuscule portion of the sample. Additionally, the interpretation of complex samples can be difficult due to the diversity of compositions both in the mixture and within a particle. In this work, we demonstrate a method to compute and cluster similarity graphs that describe inter-particle relationships within a sample using a multi-modal few-shot learning neural network. Here, as a proof-of-concept, we show that samples known to have been exposed to gunshot residue can be distinguished from samples occasionally mistaken for gunshot residue. Our workflow builds upon standard APA techniques and data processing methods to unveil additional information in a readily interpretable and quantitatively comparable format.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Hybrid Cyber-attack Detection in Photovoltaic Farms

Here, to address the cyber-physical security in PV farms, a hybrid cyber-attack detection is proposed in this manuscript. To secure PV farms, the proposed method integrates model-based and data-driven methods by fusing the detection score at the device and system levels. First, a model-based cyber-attack detection method is developed for each PV inverter. A residual between the estimation of the Kalman filter and measurement is calculated. By leveraging the calculated residual from all inverters, a squared Mahalanobis distance is developed for device detection score generation. At the system level, a convolutional neural network (CNN) is proposed to detect cyber-attack using the waveform data at the point of common coupling (PCC) in PV farms. To improve the CNN detection accuracy, a set of well-designed features are extracted from the raw waveform data. Finally, a weighted detection score fusion method is proposed to combine device and system detection scores by using their complementary strength. The feasibility and robustness of the proposed method are validated by testing cases and a comparative experiment.

14 SOLAR ENERGY↗
DOE OSTI2023

B56$δ$ long-disordered arms form a dynamic PP2A regulation interface coupled with global allostery and Jordan’s syndrome mutations

Intrinsically disordered regions (IDR) and short linear motifs (SLiMs) play pivotal roles in the intricate signaling networks governed by phosphatases and kinases. B56δ (encoded by PPP2R5D) is a regulatory subunit of protein phosphatase 2A (PP2A) with long IDRs that harbor a substrate-mimicking SLiM and multiple phosphorylation sites. De novo missense mutations in PPP2R5D cause intellectual disabilities (ID), macrocephaly, Parkinsonism, and a broad range of neurological symptoms. Our single-particle cryo-EM structures of the PP2A-B56δ holoenzyme reveal that the long, disordered arms at the B56δ termini fold against each other and the holoenzyme core. This architecture suppresses both the phosphatase active site and the substrate-binding protein groove, thereby stabilizing the enzyme in a closed latent form with dual autoinhibition. The resulting interface spans over 190 Å and harbors unfavorable contacts, activation phosphorylation sites, and nearly all residues with ID-associated mutations. Our studies suggest that this dynamic interface is coupled to an allosteric network responsive to phosphorylation and altered globally by mutations. Furthermore, we found that ID mutations increase the holoenzyme activity and perturb the phosphorylation rates, and the severe variants significantly increase the mitotic duration and error rates compared to the normal variant.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Insights into the Oxidation Mechanism of Vivianite to Metavivianite from First-Principles Calculations

Vivianite, a hydrous ferrous iron-phosphate mineral (Fe 3 (PO 4 ) 2 ·8H 2 O), readily oxidizes in contact with air yielding the less hydrous mixed-valent iron-phosphate mineral metavivianite. This topotactic transformation nominally occurs by oxidative dehydrogenation in which outgoing electrons from the iron sublattice are charge compensated by hydrolysis of structural water. However, the details of this internal charge balancing mechanism that allows the structure to remain electrostatically stable remain unknown. Here, in this study, we use density functional theory (DFT) calculations and ab initio thermodynamics (AIT) to evaluate the energetics of this process in terms of hydrogen release as a function of environmental variables such as partial pressure and temperature. The results show a thermodynamic driving force for vivianite phase transformation as oxidation progresses that is triggered by its rigid structure that has a limited accommodation for hydrogen vacancies. In contrast, metavivianite has a more flexible lattice and hydrogen bond network that stabilizes these hydrogen defects. Metavivianite is shown to be a stable intermediate for 66% residual Fe 2+ down to 33% Fe 2+ , below which other phases/structures such as santabarbarite should be more thermodynamically favorable. Our study provides a basis for experimental tests of our mechanistic findings and helps fill a basic knowledge gap about the solid-state process that defines how vivianite interacts with its surrounding environment.

Sassi, Michel [Pacific Northwest National Laborato↗
DOE OSTI2024

p14 ARF forms meso-scale assemblies upon phase separation with NPM1

NPM1 is an abundant nucleolar chaperone that, in addition to facilitating ribosome biogenesis, contributes to nucleolar stress responses and tumor suppression through its regulation of the p14 Alternative Reading Frame tumor suppressor protein (p14 ARF ). Oncogenic stress induces p14 ARF to inhibit MDM2, stabilize p53 and arrest the cell cycle. Under non-stress conditions, NPM1 stabilizes p14 ARF in nucleoli, preventing its degradation and blocking p53 activation. However, the mechanisms underlying the regulation of p14 ARF by NPM1 are unclear because the structural features of the p14 ARF -NPM1 complex were elusive. Here we show that p14 ARF assembles into a gel-like meso-scale network upon phase separation with NPM1. This assembly is mediated by intermolecular contacts formed by hydrophobic residues in an α-helix and β-strands within a partially folded N-terminal portion of p14 ARF . These hydrophobic interactions promote phase separation with NPM1, enhance p14 ARF nucleolar partitioning, restrict NPM1 diffusion within condensates and nucleoli, and reduce cellular proliferation. Our structural analysis provides insights into the multifaceted chaperone function of NPM1 in nucleoli by mechanistically linking the nucleolar localization of p14 ARF to its partial folding and meso-scale assembly upon phase separation with NPM1.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Structural mechanisms for VMAT2 inhibition by tetrabenazine

The vesicular monoamine transporter 2 (VMAT2) is a proton-dependent antiporter responsible for loading monoamine neurotransmitters into synaptic vesicles. Dysregulation of VMAT2 can lead to several neuropsychiatric disorders including Parkinson’s disease and schizophrenia. Furthermore, drugs such as amphetamine and MDMA are known to act on VMAT2, exemplifying its role in the mechanisms of actions for drugs of abuse. Despite VMAT2’s importance, there remains a critical lack of mechanistic understanding, largely driven by a lack of structural information. Here, we report a 3.1 Å resolution cryo-electron microscopy (cryo-EM) structure of VMAT2 complexed with tetrabenazine (TBZ), a non-competitive inhibitor used in the treatment of Huntington’s chorea. We find TBZ interacts with residues in a central binding site, locking VMAT2 in an occluded conformation and providing a mechanistic basis for non-competitive inhibition. We further identify residues critical for cytosolic and lumenal gating, including a cluster of hydrophobic residues which are involved in a lumenal gating strategy. Our structure also highlights three distinct polar networks that may determine VMAT2 conformational dynamics and play a role in proton transduction. The structure elucidates mechanisms of VMAT2 inhibition and transport, providing insights into VMAT2 architecture, function, and the design of small-molecule therapeutics.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Characterizing the Evolution of Trapped scCO 2 Curvature in Bentheimer and Nugget Sandstone Pore Geometry

During a Geologic Carbon Storage process, supercritical CO 2 (scCO 2 ) is subjected to a series of dynamic and static conditions where the relationship between pore geometry and the trapped scCO 2 curvature remains to be established. To mimic the dynamic process, two sandstones, Bentheimer and Nugget, were subjected to two successive drainage and imbibition (D-I) cycles and X-ray computed tomography scanned at each residual state to capture the wettability evolution at static conditions in the same pore geometry. Both sandstones contain similar grain size distributions, pore size distributions, and pore interconnectivity but differ in that the Nugget formation contains approximately half the porosity of the Bentheimer sandstone, and the pore network contains dead-end pores. scCO 2 size distributions, strain calculations, and geometric contact angle measurements were used to characterize the curvature of scCO 2 in different pore types between cycles. An increase in geometric contact angle was the greatest when advancement along the pore network of the same ganglion occurred between cycles while strain increased the most with pore-filling trapping. Moreover, Nugget sandstone results in a greater aggregated residual saturation and shows a clear increase in scCO 2 sizes with an additional D-I cycle while scCO 2 in the Bentheimer core shows a more complex response with some ganglion increasing and some decreasing in size with an additional D-I cycle. From this work, we suspect the pore geometry is playing a role in scCO 2 size distributions and use this information to suggest using water pulses to enhance trapping capacity in lower porosity sandstones.

58 GEOSCIENCES↗
DOE OSTI2020

Phylogenetics-based identification and characterization of a superior 2,3-butanediol dehydrogenase for Zymomonas mobilis expression

Abstract Background Zymomonas mobilis has recently been shown to be capable of producing the valuable platform biochemical, 2,3-butanediol (2,3-BDO). Despite this capability, the production of high titers of 2,3-BDO is restricted by several physiological parameters. One such bottleneck involves the conversion of acetoin to 2,3-BDO, a step catalyzed by 2,3-butanediol dehydrogenase (Bdh). Several Bdh enzymes have been successfully expressed in Z. mobilis, although a highly active enzyme is yet to be identified for expression in this host. Here, we report the application of a phylogenetic approach to identify and characterize a superior Bdh, followed by validation of its structural attributes using a mutagenesis approach. Results Of the 11 distinct bdh genes that were expressed in Z. mobilis, crude extracts expressing Serratia marcescens Bdh ( Sm Bdh) were found to have the highest activity (8.89 µmol/min/mg), when compared to other Bdh enzymes (0.34–2.87 µmol/min/mg). The Sm Bdh crystal structure was determined through crystallization with cofactor (NAD + ) and substrate (acetoin) molecules bound in the active site. Active Sm Bdh was shown to be a tetramer with the active site populated by a Gln247 residue contributed by the diagonally opposite subunit. Sm Bdh showed a more extensive supporting hydrogen-bond network in comparison to the other well-studied Bdh enzymes, which enables improved substrate positioning and substrate specificity. This protein also contains a short α6 helix, which provides more efficient entry and exit of molecules from the active site, thereby contributing to enhanced substrate turnover. Extending the α6 helix to mimic the lower activity Enterobacter cloacae ( Ec Bdh) enzyme resulted in reduction of Sm Bdh function to nearly 3% of the total activity. In great contrast, reduction of the corresponding α6 helix of the Ec Bdh to mimic the Sm Bdh structure resulted in ~ 70% increase in its activity. Conclusions This study has demonstrated that Sm Bdh is superior to other Bdhs for expression in Z. mobilis for 2,3-BDO production. Sm Bdh possesses unique structural features that confer biochemical advantage to this protein. While coordinated active site formation is a unique structural characteristic of this tetrameric complex, the smaller α6 helix and extended hydrogen network contribute towards improved activity and substrate promiscuity of the enzyme.

09 BIOMASS FUELS↗
DOE OSTI2022

Excellent antioxidizing, thermally insulating and flame resistance silica‐polybenzoxazine aerogels for aircraft ablative materials

Abstract High‐performance thermal protective composites with lightweight, micro‐ ablation and high‐efficient thermal insulation are urgently required for thermal protection systems in advanced hypersonic speed vehicles. However, the practical applications of thermal protective composites have long been hampered by the main issues such as low mass residual rate and poor long‐term antioxidation of the matrix in high‐temperature aerobic environments. Here, we report a novel silica‐polybenz oxazine (SiO 2 ‐PBO) aerogels with interpenetrated networks, possessing the ability to antioxidation, thermal insulation, and flame‐retardant properties. The resulting SiO 2 ‐PBO aerogels exhibit low density (0.25 g/cm 3 ), low thermal conductivity (0.035 W/(m·K)), and superior peak heat release rate value (15.3 W/g). Moreover, the mass residual rate is up to 70.46 wt% in the N 2 atmosphere and remains 57.83 wt% despite existing in the air atmosphere and experiencing the highest temperature of 800°C. Briefly, SiO 2 ‐PBO aerogels as‐prepared could be a potential matrix for a new gene ration of high‐performance thermal protective composites in the future.

Xiao, Yunyun↗
DOE OSTI2023

An integrated coupling model for solving multiscale fluid-fluid coupling problems in SAM code

In this study, an integrated coupling method has been developed for solving multiscale fluid-fluid coupling problems in plant-scale safety analysis models in SAM (System Analysis Module) code. In this method, a higher-fidelity multi-dimensional (3D) flow module is used for reactor components of complex flow features (e.g., reactor core) and a lumped parameter one-dimensional (1D) flow module for plant-scale flow loops (e.g., primary loop pipe network), respectively. In this method, the 3D fluid equation/domain and 1D fluid equation/domain are tightly coupled at the residual level and solved simultaneously using the Newton’s method to overcome the convergence issues typically seen in existing approaches like separate domain approach, where the 3D fluid equation and 1D fluid equation are solved separately. Extensive and successful code verifications and demonstrations have been performed for this newly developed method. This new modeling approach significantly simplify the work flow in developing high-fidelity plant-scale safety analysis model, e.g. for pool-type reactors and pebble-bed reactors.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2025

Reducing NOx Emissions in Ammonia Combustors

Ammonia continues to attract growing interest as a carbonneutral replacement fuel, motivating numerous research efforts toward understanding fundamental ammonia combustion characteristics. A major challenge for the use of ammonia is the development of combustor technologies for mitigating potentially high NOx emissions from the fuel-bound nitrogen chemical pathways to acceptable levels. Our work focuses on a staged RQL combustor architecture for minimizing the NOx emission levels through burning fuel-rich in the primary stage to formcombustion products containing significant levels of hydrogen in addition to nitrogen and water with minimal NOx formation. The subsequent quench and burnout stages of the combustor must then quickly burn residual hydrogen with flame-temperatures moderated by nitrogen and water forming in the first stage. Chemical Reactor Network modeling was used to understand and identify optimal stoichiometry and residence times in each stage for minimizing NOx emissions and to quantify pressure and temperature effects. Reducing the overall NOx emissions requires relatively long residence times in the primary stage to achieve near equilibrium NO levels due to kinetically controlling processes. For conditions relevant to gas turbines (e.g., 30 atm), our work indicates that NOx emissions below 20 ppm are theoretically achievable in a staged RQL combustor architecture. However, these emission predictions significantly depend on the accuracies of currently available chemical kinetic mechanisms which have not been extensively validated under elevated pressure and temperature conditions relevant to gas turbines.

10 SYNTHETIC FUELS↗
DOE OSTI2025

Chemical Reactor Network Modeling of Ammonia Rich-Quench-Lean Combustion Using a Partially Stirred Reactor Approach

Ammonia is a promising alternative to hydrogen with high energy density and favorable storage and transport characteristics. However, low flammability and a propensity for high nitrogen oxide (NO x ) emissions make direct utilization challenging. Recently, two-stage rich-quench-lean (RQL) combustion strategies have shown promise in achieving low NO x emissions with ammonia. In this approach, the rich stage serves to oxidize a portion of the fuel while thermally decomposing as much of the remaining ammonia as possible, generating hydrogen. In the second (lean) stage, air is rapidly introduced, burning out the hydrogen and residual ammonia. Two-stage RQL combustion of ammonia has been investigated in the open literature both experimentally and numerically. In general, idealized chemical reactor network (CRN) models predict NO x concentrations below those of 2D/3D computational fluid dynamics models and experiments. The primary drivers of these discrepancies may be largely attributed to finite rate mixing nonadiabatic operation. The typical CRN model is comprised of a perfectly-stirred-reactor (PSR), followed by a plug-flow-reactor (PFR), meant to represent the flame, and postflame zones, respectively. In the two-stage RQL approach two PSR-PFR networks are arranged sequentially, corresponding to the rich and lean stages, with secondary air injection in between. In the authors' past work, this arrangement has demonstrated the significant sensitivity of exit NO x to the rich stage equivalence ratio, while the amount of secondary air injection was shown to be less critical. In this paper, the CRN model is extended to (1) include the impacts of heat loss and (2) utilize a partially-stirred-reactor (PaSR) approach to study the impacts of mixing on emissions performance. Varying amounts of heat loss are applied to the rich relaxation zone to understand emissions performance and changes to optimization of equivalence ratio and residence time. Premixed and nonpremixed configurations are considered in the rich stage PaSR, with varying degrees of mixing intensity to study the interaction between mixing, transport, and kinetic timescales. Critically, the impact of mixing between hot products and secondary air injection is studied to understand practical injector needs. Results show unburnt ammonia leaving the rich stage as a primary contributor to NO x emissions – driven both by increased heat loss and reduced mixing rates. Furthermore, heat losses have been shown to create conditions that are conducive to increased N 2 O formation in the lean stage. In conclusion, the results of this study will be considered in the context of developing optimized two-stage RQL combustors for ammonia.

Combustion↗
DOE OSTI2025

Three-Dimensional Pore Networks in Miocene Stevens Sandstone of California: Implications for CO 2 Geologic Storage

The Miocene Stevens Sandstone in the San Joaquin Basin of California is increasingly recognized as a promising candidate for CO 2 geological storage due to the enormous storage capacity, proven sealing, and existing infrastructure. In this study, computed microtomography imaging and pore network modeling were employed to investigate the influence of pore geometry and wettability on the CO 2 injectivity and residual trapping. Image analysis revealed that a significant fraction of the cement and matrix consists of microporous regions. The microporosity can substantially increase the overall pore space, yet its contribution to permeability remains modest, particularly in samples with low permeability. The intrinsic heterogeneity of turbidite reservoirs further complicates the reservoir properties among different layers. Two-phase flow simulations under varying wettability conditions (water-wet, weak water-wet, and neutral-wet) demonstrated that the CO 2 injection is predominantly controlled by macropores. CO 2 invades microporous regions only after these larger pores are filled. The presence of microporosity leads to a decrease in both initial and residual CO 2 saturations, with the magnitude of the reduction being influenced by wettability. Neutral-wet scenarios exhibit higher CO 2 mobility and thus lower residual trapping than water-wet scenarios. The results imply that heterogeneity in pore geometry and cement distribution across different layers can result in stratified CO 2 flow pathways, complicating efforts to predict injection performance. Overall, the Stevens Sandstone shows considerable promise for CO 2 geologic storage, but effective implementation will require detailed characterization of the pore structure as well as the integration of reactive fluid flow to account for potential mineral dissolution and fines migration.

fluids↗
DOE OSTI2021

LX-17 Thermal Decomposition-Characterization of Solid Residues from Cook-Off in a Small-Scale Vessel Under Confinement

Concerns surround whether insensitive (or any) energetic materials are more dangerous to handle when exposed to abnormal thermal environments. This study characterizes the residual material remaining after LX-17 (92.5 % 1,3,5-triamino 2,4,6-trinitro benzene (TATB) and 7.5 % Kel-F) is exposed to various thermal environments in a sealed small-scale vessel cook-off test reactor (heated at 0.1 to 100 °C/min until the reactor opened at 3000 psi (20.7 MPa)). Previous work has shown no additional sensitivity of these residues as evaluated by small-scale safety analysis, but characterization on the molecular scale indicates the TATB is transformed to more reactive compounds as well as the residue could be precursors to toxic gases. The solids and chars were characterized by various analytical methods. Heat-flow measurements indicated exothermic release is due to a mixture of residual TATB and related decomposition products (which may be more energetic). The N/C and O/N ratios indicated a material much more degraded than TATB. Primarily, the solids were a network of amorphous C inter-dispersed with N and O. Types of bonding include C–C, C–N, N–H, N–C, N=C, N≡C, C–O=, and –OH. Solvent extracts of the solids showed TATB decomposition intermediates benzo-furazans and benzo-furoxans, substituted TATB (mono-nitroso, hydroxyl, and chlorinated) along with several unidentified smaller molecules. These results indicate thermal treatment produces an amorphous carbon residue with heteroatoms incorporated through differing functionality, varying depending upon the thermal severity of exposure. These structures also could further decompose producing toxic light gases (such as cyanide).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Primordial non-Gaussianity from the completed SDSS-IV extended Baryon Oscillation Spectroscopic Survey – I: Catalogue preparation and systematic mitigation

ABSTRACT We investigate the large-scale clustering of the final spectroscopic sample of quasars from the recently completed extended Baryon Oscillation Spectroscopic Survey (eBOSS). The sample contains 343 708 objects in the redshift range 0.8 < z < 2.2 and 72 667 objects with redshifts 2.2 < z < 3.5, covering an effective area of $4699\, {\rm deg}^{2}$. We develop a neural network-based approach to mitigate spurious fluctuations in the density field caused by spatial variations in the quality of the imaging data used to select targets for follow-up spectroscopy. Simulations are used with the same angular and radial distributions as the real data to estimate covariance matrices, perform error analyses, and assess residual systematic uncertainties. We measure the mean density contrast and cross-correlations of the eBOSS quasars against maps of potential sources of imaging systematics to address algorithm effectiveness, finding that the neural network-based approach outperforms standard linear regression. Stellar density is one of the most important sources of spurious fluctuations, and a new template constructed using data from the Gaia spacecraft provides the best match to the observed quasar clustering. The end-product from this work is a new value-added quasar catalogue with the improved weights to correct for non-linear imaging systematic effects, which will be made public. Our quasar catalogue is used to measure the local-type primordial non-Gaussianity in a companion paper.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Direct Bayesian inference for fault severity assessment in Digital-Twin-Based fault diagnosis

For applications in condition-based maintenance of nuclear systems, the assessment of fault severity is crucial. In this work, we developed a framework that allows for direct inference of the probability distributions of possible faults in a system. Here, we employed a model-based approach with model residuals generated from analytical redundancy relations provided by physics-based models of the system components. From real-time sensor readings, the values of the model residuals can be calculated, and the posterior probability distributions of the faults can be computed directly using the methods of Bayesian networks. From the posterior distribution of each fault, one can estimate the fault probability based on a chosen threshold and assess the severity of the fault. By eliminating the discretization and simplifications in middle steps, this approach allows us to leverage the available computational resources to provide more accurate fault probability estimates and severity assessments.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗