Search NASA⌕ Search

SEARCH · Search NASA

Results for “Model 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 1,063 records · Page 59
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Impact of Limited Degree of Freedom Drag Coefficients on a Floating Offshore Wind Turbine Simulation

The worldwide effort to design and commission floating offshore wind turbines (FOWT) is motivating the need for reliable numerical models that adequately represent their physical behavior under realistic sea states. However, properly representing the hydrodynamic quadratic damping for FOWT remains uncertain, because of its dependency on the choice of drag coefficients (dimensionless or not). It is hypothesized that the limited degree of freedom (DoF) drag coefficient formulation that uses only translational drag coefficients causes mischaracterization of the rotational DoF drag, leading to underestimation of FOWT global loads, such as tower base fore-aft shear. To address these hydrodynamic modeling uncertainties, different quadratic drag models implemented in the open-source mid-fidelity simulation tool, OpenFAST, were investigated and compared with the experimental data from the Offshore Code Comparison Collaboration, Continued, with Correlation (OC5) project. The tower base fore-aft shear and up-wave mooring line tension were compared under an irregular wave loading condition to demonstrate the effects of the different damping models. Two types of hydrodynamic quadratic drag formulations were considered: (1) member-based dimensionless drag coefficients applied only at the translational DoF (namely limited-DoF drag model) and (2) quadratic drag matrix model (in dimensional form). Based on the results, the former consistently underestimated the 95th percentile peak loads and spectral responses when compared to the OC5 experimental data. In contrast, the drag matrix models reduced errors in estimates of the tower base shear peak load by 7–10% compared to the limited-DoF drag model. The underestimation in the tower base fore-aft shear was thus inferred be related to mischaracterization of the rotational pitch drag and the heave motion/drag by the limited-DoF model.

17 WIND ENERGY↗
DOE OSTI2020

Measurement and modeling of the radio frequency sheath impedance in a large magnetized plasma

The DC and RF properties of radio frequency (RF) driven sheaths were studied in the Large Plasma Device (LAPD) at the University of California, Los Angeles. The experiments diagnosed RF sheaths on field lines connected to a grounded plate at one end and an ion cyclotron range of frequencies (ICRF) antenna at the other end. The experimental setup permitted measurement of the RF sheath impedance at the plate as a function of DC sheath voltage, with the latter controlled by varying the RF current applied to the antenna. Here, the DC current-voltage characteristics of these sheaths and the RF sheath impedance measurements were compared with modeling. Hot electrons, present in the LAPD plasma, were inferred to contribute significantly to both the DC and RF currents and hence the RF impedance. It was postulated that at very low power hot electrons could not access the region of the plasma subject to RF waves, resulting in an increased RF impedance. Within some experimental limitations and significant assumptions, an RF sheath impedance model was verified by the experimental data.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Mapping Stellar Surfaces. II. An Interpretable Gaussian Process Model for Light Curves

The use of Gaussian processes (GPs) as models for astronomical time series data sets has recently become almost ubiquitous, given their ease of use and flexibility. In particular, GPs excel at marginalization over the stellar signal when the variability due to starspots is treated as a nuisance, as in exoplanet transit modeling. However, these effective models are less useful in cases where the starspot signal is of primary interest, since it is not obvious how the parameters of the GP relate to physical parameters like the spot size, contrast, and latitudinal distribution. Instead, it is common practice to explicitly model the effect of individual starspots on the light curve and attempt to infer their properties via optimization or posterior inference. Unfortunately, this process is ill-posed and often computationally intractable when applied to stars with more than a few spots and/or to ensembles of many stars. Here we derive a closed-form expression for a GP that describes the light curve of a rotating, evolving stellar surface conditioned on a given distribution of starspot sizes, contrasts, and latitudes. We demonstrate that this model is correctly calibrated, allowing one to robustly infer physical parameters of interest from one or more light curves, including the typical spot radii and latitudes. Our GP has far-ranging implications for understanding the variability and magnetic activity of stars from light curves and radial velocity measurements, as well as for modeling correlated noise in exoplanet searches. Our implementation is efficient, user-friendly, and open-source, available in the package starry-process.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Using a surrogate-assisted Bayesian framework to calibrate the runoff-generation scheme in the Energy Exascale Earth System Model (E3SM) v1

Abstract. Runoff is a critical component of the terrestrial water cycle, and Earth system models (ESMs) are essential tools to study its spatiotemporal variability. Runoff schemes in ESMs typically include many parameters so that model calibration is necessary to improve the accuracy of simulated runoff. However, runoff calibration at a global scale is challenging because of the high computational cost and the lack of reliable observational datasets. In this study, we calibrated 11 runoff relevant parameters in the Energy Exascale Earth System Model (E3SM) Land Model (ELM) using a surrogate-assisted Bayesian framework. First, the polynomial chaos expansion machinery with Bayesian compressed sensing is used to construct computationally inexpensive surrogate models for ELM-simulated runoff at 0.5∘ × 0.5∘ for 1991–2010. The error metric between the ELM simulations and the benchmark data is selected to construct the surrogates, which facilitates efficient calibration and avoids the more conventional, but challenging, construction of high-dimensional surrogates for the ELM simulated runoff. Second, the Sobol' index sensitivity analysis is performed using the surrogate models to identify the most sensitive parameters, and our results show that, in most regions, ELM-simulated runoff is strongly sensitive to 3 of the 11 uncertain parameters. Third, a Bayesian method is used to infer the optimal values of the most sensitive parameters using an observation-based global runoff dataset as the benchmark. Our results show that model performance is significantly improved with the inferred parameter values. Although the parametric uncertainty of simulated runoff is reduced after the parameter inference, it remains comparable to the multimodel ensemble uncertainty represented by the global hydrological models in ISMIP2a. Additionally, the annual global runoff trend during the simulation period is not well constrained by the inferred parameter values, suggesting the importance of including parametric uncertainty in future runoff projections.

58 GEOSCIENCES↗
DOE OSTI2020

Evidence and modeling of turbulence bifurcation in L-mode confinement transitions on Alcator C-Mod

Analysis and modeling of rotation reversal hysteresis experiments show that a single turbulent bifurcation is responsible for the Linear to Saturated Ohmic Confinement (LOC/SOC) transition and concomitant intrinsic rotation reversal on Alcator C-Mod. Plasmas on either side of the reversal exhibit different toroidal rotation profiles and therefore different turbulence characteristics despite the profiles of density and temperature, which are indistinguishable within measurement uncertainty. Elements of this bifurcation are also shown to persist for auxiliary heated L-modes. The deactivation of subdominant (in the linear growth rate and contribution to heat transport) ion temperature gradient and trapped electron mode instabilities is identified as the only possible change in turbulence within a reduced quasilinear transport model across the reversal, which is consistent with the measured profiles and inferred heat and particle fluxes. Experimental constraints on a possible change from strong to weak turbulence, outside the description of the quasilinear model, are also discussed. These results indicate an explanation for the LOC/SOC transition that provides a mechanism for the hysteresis through the dynamics of subdominant modes and changes in their relative populations and does not involve a change in the most linearly unstable ion-scale drift-wave instability.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Early retirement of power plants in climate mitigation scenarios

International efforts to avoid dangerous climate change aim for large and rapid reductions of fossil fuel CO 2 emissions worldwide, including nearly complete decarbonization of the electric power sector. However, achieving such rapid reductions may depend on early retirement of coal- and natural gas-fired power plants. Here, we analyze future fossil fuel electricity demand in 171 energy-emissions scenarios from Integrated Assessment Models (IAMs), evaluating the implicit retirements and/or reduced operation of generating infrastructure. Although IAMs calculate retirements endogenously, the structure and methods of each model differ; we use a standard approach to infer retirements in outputs from all six major IAMs and—unlike the IAMs themselves—we begin with the age distribution and region-specific operating capacities of the existing power fleet. We find that coal-fired power plants in scenarios consistent with international climate targets (i.e. keeping global warming well-below 2 °C or 1.5 °C) retire one to three decades earlier than historically has been the case. If plants are built to meet projected fossil electricity demand and instead allowed to operate at the level and over the lifetimes they have historically, the roughly 200 Gt CO 2 of additional emissions this century would be incompatible with keeping global warming well-below 2 °C. Thus, ambitious climate mitigation scenarios entail drastic, and perhaps un-appreciated, changes in the operating and/or retirement schedules of power infrastructure.

20 FOSSIL-FUELED POWER PLANTS↗
DOE OSTI2023

Advanced data analysis in inertial confinement fusion and high energy density physics

Bayesian analysis enables flexible and rigorous definition of statistical model assumptions with well-characterized propagation of uncertainties and resulting inferences for single-shot, repeated, or even cross-platform data. This approach has a strong history of application to a variety of problems in physical sciences ranging from inference of particle mass from multi-source high-energy particle data to analysis of black-hole characteristics from gravitational wave observations. The recent adoption of Bayesian statistics for analysis and design of high-energy density physics (HEDP) and inertial confinement fusion (ICF) experiments has provided invaluable gains in expert understanding and experiment performance. In this Review, we discuss the basic theory and practical application of the Bayesian statistics framework. We highlight a variety of studies from the HEDP and ICF literature, demonstrating the power of this technique. Due to the computational complexity of multi-physics models needed to analyze HEDP and ICF experiments, Bayesian inference is often not computationally tractable. Two sections are devoted to a review of statistical approximations, efficient inference algorithms, and data-driven methods, such as deep-learning and dimensionality reduction, which play a significant role in enabling use of the Bayesian framework. We provide additional discussion of various applications of Bayesian and machine learning methods that appear to be sparse in the HEDP and ICF literature constituting possible next steps for the community. We conclude by highlighting community needs, the resolution of which will improve trust in data-driven methods that have proven critical for accelerating the design and discovery cycle in many application areas.

47 OTHER INSTRUMENTATION↗
DOE OSTI2026

Smokescreen: A Python package for data vector blinding and encryption in cosmological analyses

Smokescreen is an open-source Python library for data-vector concealment (blinding) in cosmological analyses. Data-vector blinding works by applying cosmology-dependent shifts to the observed data vector, moving it away from the true cosmological signal without affecting its statistical properties, so that analysts cannot infer the true result until the analysis is frozen and the blinding is lifted. The package computes these shifts using Firecrown likelihoods applied to data vectors stored in the SACC format, ensuring that the theoretical model used for blinding is identical to that used for inference whilst remaining agnostic to the specific observable being blinded. To prevent accidental unblinding, the original SACC file, containing the true cosmology, is encrypted. Although developed for the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), Smokescreen is applicable to any experiment using Firecrown likelihoods and the SACC data format.

Loureiro, Arthur [Stockholm U., OKC; Imperial Coll↗
DOE OSTI2024

The Modeled Seasonal Cycles of Surface N 2 O Fluxes and Atmospheric N 2 O

Nitrous oxide (N 2 O) is a greenhouse gas and stratospheric ozone-depleting substance with large and growing anthropogenic emissions. Previous studies identified the influx of N 2 O-depleted air from the stratosphere to partly cause the seasonality in tropospheric N 2 O (aN 2 O), but other contributions remain unclear. Here, we combine surface fluxes from eight land and four ocean models from phase 2 of the Nitrogen/N 2 O Model Intercomparison Project with tropospheric transport modeling to simulate aN 2 O at eight remote air sampling sites for modern and pre-industrial periods. Models show general agreement on the seasonal phasing of zonal-average N 2 O fluxes for most sites, but seasonal peak-to-peak amplitudes differ several-fold across models. The modeled seasonal amplitude of surface aN 2 O ranges from 0.25 to 0.80 ppb (interquartile ranges 21%–52% of median) for land, 0.14–0.25 ppb (17%–68%) for ocean, and 0.28–0.77 ppb (23%–52%) for combined flux contributions. The observed seasonal amplitude ranges from 0.34 to 1.08 ppb for these sites. The stratospheric contributions to aN 2 O, inferred by the difference between the surface-troposphere model and observations, show 16%–126% larger amplitudes and minima delayed by ~1 month compared to Northern Hemisphere site observations. Land fluxes and their seasonal amplitude have increased since the pre-industrial era and are projected to grow further under anthropogenic activities. Our results demonstrate the increasing importance of land fluxes for aN 2 O seasonality. Considering the large model spread, in situ aN 2 O observations and atmospheric transport-chemistry models will provide opportunities for constraining terrestrial and oceanic biosphere models, critical for projecting carbon-nitrogen cycles under ongoing global warming.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Statistical tests of simple earthquake cycle models: Testing Earthquake Cycle Models

A central goal of observing and modeling the earthquake cycle is to forecast when a particular fault may generate an earthquake: a fault late in its earthquake cycle may be more likely to generate an earthquake than a fault early in its earthquake cycle. Models that can explain geodetic observations throughout the entire earthquake cycle may be required to gain a more complete understanding of relevant physics and phenomenology. Previous efforts to develop unified earthquake models for strike-slip faults have largely focused on explaining both preseismic and postseismic geodetic observations available across a few faults in California, Turkey, and Tibet. An alternative approach leverages the global distribution of geodetic and geologic slip rate estimates on strike-slip faults worldwide. Here we use the Kolmogorov-Smirnov test for similarity of distributions to infer, in a statistically rigorous manner, viscoelastic earthquake cycle models that are inconsistent with 15 sets of observations across major strike-slip faults. We reject a large subset of two-layer models incorporating Burgers rheologies at a significance level of α = 0.05 (those with long-term Maxwell viscosities η M <~ 4.0 × 10 19 Pa s and η M >~ 4.6 × 10 20 Pa s) but cannot reject models on the basis of transient Kelvin viscosity η K . Finally, we examine the implications of these results for the predicted earthquake cycle timing of the 15 faults considered and compare these predictions to the geologic and historical record.

58 GEOSCIENCES↗
DOE OSTI2024

Learning of networked spreading models from noisy and incomplete data

Recent years have seen a lot of progress in algorithms for learning parameters of spreading dynamics from both full and partial data. Some of the remaining challenges include model selection under the scenarios of unknown network structure, noisy data, missing observations in time, as well as an efficient incorporation of prior information to minimize the number of samples required for an accurate learning. Here, in this work, we introduce a universal learning method based on a scalable dynamic message-passing technique that addresses these challenges often encountered in real data. The algorithm leverages available prior knowledge on the model and on the data, and reconstructs both network structure and parameters of a spreading model. We show that a linear computational complexity of the method with the key model parameters makes the algorithm scalable to large network instances.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Population-level Dark Energy Constraints from Strong Gravitational Lensing using Simulation-Based Inference

In this work, we present a scalable approach for inferring the dark energy equation-of-state parameter ($w$) from a population of strong gravitational lens images using Simulation-Based Inference (SBI). Strong gravitational lensing offers crucial insights into cosmology, but traditional Monte Carlo methods for cosmological inference are computationally prohibitive and inadequate for processing the thousands of lenses anticipated from future cosmic surveys. New tools for inference, such as SBI using Neural Ratio Estimation (NRE), address this challenge effectively. By training a machine learning model on simulated data of strong lenses, we can learn the likelihood-to-evidence ratio for robust inference. Our scalable approach enables more constrained population-level inference of $w$ compared to individual lens analysis, constraining $w$ to within $1\sigma$. Our model can be used to provide cosmological constraints from forthcoming strong lens surveys, such as the 4MOST Strong Lensing Spectroscopic Legacy Survey (4SLSLS), which is expected to observe 10,000 strong lenses.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

Population-level Dark Energy Constraints from Strong Gravitational Lensing using Simulation-Based Inference

In this work, we present a scalable approach for inferring the dark energy equation-of-state parameter ($w$) from a population of strong gravitational lens images using Simulation-Based Inference (SBI). Strong gravitational lensing offers crucial insights into cosmology, but traditional Monte Carlo methods for cosmological inference are computationally prohibitive and inadequate for processing the thousands of lenses anticipated from future cosmic surveys. New tools for inference, such as SBI using Neural Ratio Estimation (NRE), address this challenge effectively. By training a machine learning model on simulated data of strong lenses, we can learn the likelihood-to-evidence ratio for robust inference. Our scalable approach enables more constrained population-level inference of $w$ compared to individual lens analysis, constraining $w$ to within 1$\sigma$. Our model can be used to provide cosmological constraints from forthcoming strong lens surveys, such as the 4MOST Strong Lensing Spectroscopic Legacy Survey (4SLSLS), which is expected to observe 10,000 strong lenses.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2018

Extreme-Scale Bayesian Inference for Uncertainty Quantification of Complex Simulations

Uncertainty quantification (UQ)—that is, quantifying uncertainties in complex mathematical models and their large-scale computational implementations—is widely viewed as one of the outstanding challenges facing the field of CS&E over the coming decade. The EUREKA project set to address the most difficult class of UQ problems: those for which both the underlying PDE model as well as the uncertain parameters are of extreme scale. In the project we worked on these extreme-scale challenges in the following four areas: 1. Scalable parallel algorithms for sampling and characterizing the posterior distribution that exploit the structure of the underlying PDEs and parameter-to-observable map. These include structure-exploiting versions of the randomized maximum likelihood method, which aims to overcome the intractability of employing conventional MCMC methods for solving extreme-scale Bayesian inversion problems by appealing to and adapting ideas from large-scale PDE-constrained optimization, which have been very successful at exploring high-dimensional spaces. 2. Scalable parallel algorithms for construction of prior and likelihood functions based on learning methods and non-parametric density estimation. Constructing problem-specific priors remains a critical challenge in Bayesian inference, and more so in high dimensions. Another challenge is construction of likelihood functions that capture unmodeled couplings between observations and parameters. We will create parallel algorithms for non-parametric density estimation using high dimensional N-body methods and combine them with supervised learning techniques for the construction of priors and likelihood functions. 3. Bayesian inadequacy models, which augment physics models with stochastic models that represent their imperfections. The success of the Bayesian inference framework depends on the ability to represent the uncertainty due to imperfections of the mathematical model of the phenomena of interest. This is a central challenge in UQ, especially for large-scale models. We propose to develop the mathematical tools to address these challenges in the context of extreme-scale problems. 4. Parallel scalable algorithms for Bayesian optimal experimental design (OED). Bayesian inversion yields quantified uncertainties in the model parameters, which can be propagated forward through the model to yield uncertainty in outputs of interest. This opens the way for designing new experiments to reduce the uncertainties in the model parameters and model predictions. Such experimental design problems have been intractable for large-scale problems using conventional methods; we will create OED algorithms that exploit the structure of the PDE model and the parameter-to-output map to overcome these challenges. Parallel algorithms for these four problems were created, analyzed, prototyped, implemented, tuned, and scaled up for leading-edge supercomputers, including UT-Austin’s own 10 petaflops Stampede system, ANL’s Mira system, and ORNL’s Titan system. While our focus is on fundamental mathematical/computational methods and algorithms, we will assess our methods on model problems derived from several DOE mission applications, including multiscale mechanics and ice sheet dynamics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Signs of early-stage disk growth revealed with ALMA

We present ALMA 1.3 mm continuum, {sup 12}CO, C{sup 18}O, and SO data for the Class 0 protostars Lupus 3 MMS, IRAS 15398−3559, and IRAS 16253−2429 at resolutions of ∼100 au. By measuring a rotational profile in C{sup 18}O, a 100 au Keplerian disk around a 0.3 M{sub ⊙} protostar is observed in Lupus 3 MMS. No 100 au Keplerian disks are observed in IRAS 15398−3559 and IRAS 16253−2429. Nevertheless, embedded compact (<30 au) continuum components are detected. The C{sup 18}O emission in IRAS 15398−3559 shows signatures of infall with a constant angular momentum. IRAS 16253−2429 exhibits signatures of infall and rotation, but its rotational profile is unresolved. By fitting the C{sup 18}O data with our kinematic models, the protostellar masses and the disk radii are inferred to be 0.01 M{sub ⊙} and 20 au in IRAS 15398−3559, and 0.03 M{sub ⊙} and 6 au in IRAS 16253−2429. By comparing the specific angular momentum profiles from 10,000 au to 100 au in eight Class 0 and I protostars, we find that the evolution of envelope rotation can be described with conventional inside-out collapse models. In comparison with a sample of 18 protostars with known disk radii, our results reveal signs of disk growth, with the disk radius increasing as M{sub ∗}{sup 0.8±0.14} or t{sup 1.09±0.37} in the Class 0 stage, where M{sub *} is the protostellar mass and t is the age. The disk growth rate slows down in the Class I stage. In addition, we find a hint that the mass accretion rate declines as t{sup −0.26±0.04} from the Class 0 to the Class I stages.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

The Mean Metal-line Absorption Spectrum of Damped Ly α Systems in BOSS

We study the mean absorption spectrum of the Damped Ly α (DLA) population at z ∼ 2.6 by stacking normalized, rest-frame-shifted spectra of ∼27,000 DLA systems from the DR12 of the Baryon Oscillation Spectroscopic Survey (BOSS)/SDSS-III. We measure the equivalent widths of 50 individual metal absorption lines in five intervals of DLA hydrogen column density, five intervals of DLA redshift, and overall mean equivalent widths for an additional 13 absorption features from groups of strongly blended lines. The mean equivalent width of low-ionization lines increases with N {sub H} {sub i}, whereas for high-ionization lines the increase is much weaker. The mean metal line equivalent widths decrease by a factor ∼1.1–1.5 from z ∼ 2.1 to z ∼ 3.5, with small or no differences between low- and high-ionization species. We develop a theoretical model, inspired by the presence of multiple absorption components observed in high-resolution spectra, to infer mean metal column densities from the equivalent widths of partially saturated metal lines. We apply this model to 14 low-ionization species and to Al iii, S iii, Si iii, C iv, Si iv, N v, and O vi. We use an approximate derivation for separating the equivalent width contributions of several lines to blended absorption features, and infer mean equivalent widths and column densities from lines of the additional species N i, Zn ii, C ii*, Fe iii, and S iv. Several of these mean column densities of metal lines in DLAs are obtained for the first time; their values generally agree with measurements of individual DLAs from high-resolution, high signal-to-noise ratio spectra when they are available.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Estimation of stagnation performance metrics in magnetized liner inertial fusion experiments using Bayesian data assimilation

Here we present a new analysis methodology that allows for the self-consistent integration of multiple diagnostics including nuclear measurements, x-ray imaging, and x-ray power detectors to determine the primary stagnation parameters, such as temperature, pressure, stagnation volume, and mix fraction in magnetized liner inertial fusion (MagLIF) experiments. The analysis uses a simplified model of the stagnation plasma in conjunction with a Bayesian inference framework to determine the most probable configuration that describes the experimental observations while simultaneously revealing the principal uncertainties in the analysis. We validate the approach by using a range of tests including analytic and three-dimensional MHD models. An ensemble of MagLIF experiments is analyzed, and the generalized Lawson criterion χ is estimated for all experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

VpROM: a novel variational autoencoder-boosted reduced order model for the treatment of parametric dependencies in nonlinear systems

Reduced Order Models (ROMs) are of considerable importance in many areas of engineering in which computational time presents difficulties. Established approaches employ projection-based reduction, such as Proper Orthogonal Decomposition. The limitation of the linear nature of such operators is typically tackled via a library of local reduction subspaces, which requires the assembly of numerous local ROMs to address parametric dependencies. Our work attempts to define a more generalisable mapping between parametric inputs and reduced bases for the purpose of generative modeling. We propose the use of Variational Autoencoders (VAEs) in place of the typically utilised clustering or interpolation operations, for inferring the fundamental vectors, termed as modes, which approximate the manifold of the model response for any and each parametric input state. The derived ROM still relies on projection bases, built on the basis of full-order model simulations, thus retaining the imprinted physical connotation. However, it additionally exploits a matrix of coefficients that relates each local sample response and dynamics to the global phenomena across the parametric input domain. The VAE scheme is utilised for approximating these coefficients for any input state. This coupling leads to a high-precision low-order representation, which is particularly suited for problems where model dependencies or excitation traits cause the dynamic behavior to span multiple response regimes. Moreover, the probabilistic treatment of the VAE representation allows for uncertainty quantification on the reduction bases, which may then be propagated to the ROM response. The performance of the proposed approach is validated on an open-source simulation benchmark featuring hysteresis and multi-parametric dependencies, and on a large-scale wind turbine tower characterised by nonlinear material behavior and model uncertainty.

Conditional VAEs↗