Search NASA⌕ Search

SEARCH · Search NASA

Results for “variance reduction”

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 163 records · Page 9

GANISP (GAN-assisted Importance SPlitting)

Genealogical importance splitting marches towards a rare event by iteratively selecting and replicating realizations that are headed towards a rare event. The replication step is made difficult when applied to deterministic systems as the initial conditions of the offspring realizations need to be adjusted. Typically, a random perturbation is applied to the offspring. For some cases, this cloning technique may not be adequate and prevent variance reduction in the probability estimate. A GAN-based replication process is proposed to address this limitation. The perturbations applied to the clones are physically consistent instead of being randomly chosen. The proposed method allows reducing the variance in the probability estimation.

Hassanaly, Malik↗

Towards a Robust Sampling Approach: A Computational Review and Design

As pointed out in several other works, the estimation of the reliability of the electrical grid can not be conducted without the estimation of the stochastic phenomena of electicity demand and electricity production by variable renewable sources. Therefore, sampling procedures have become integral in the design of engineering structures and analysis. Commonly referred to as Monte Carlo uncertainty analysis or integration, the general objective of these procedures is to establish specifics about the uncertainty of an output characteristic of such an engineering system, given uncertainty about its input characteristics. These sampling procedures are applied in a context in which establishing such specifics cannot be performed through other means. Variance-reduction techniques are designed to lessen the variability among estimators to estimate statistics of those output uncertainties. Importance-sampling techniques, on the other hand, are designed to reduce the number of samples needed to estimate a particular statistic—e.g., a tail probability. The combination of these approaches can reduce the computational burden considerably for a particular estimator and statistic. Importance sampling—geared and designed as it is toward improving a particular estimator—suffers, unfortunately, from the unintended consequence of reducing the performance of other estimators in terms of their variance. The objective of this paper is to offer an alternative sampling procedure where this variance does not grow unacceptably large for a suite of estimators. Moreover, it is anticipated that, with additional knowledge of how an engineering output characteristic responds to its input characteristics, tuning parameters of the input’s sampling procedure can be set to improve the output characteristic’s estimation. This work proves the effectiveness of the suggested alternative approach. Such positive outcome will lead to a decrease of the computational burden of performing stochastic optimization of integrated energy systems (e.g., components dispatch, and portfolio composition). In particular, capturing the contribution to the overall system cost of rare and unlikely events and patterns of the electricity demand and production will become less computationally expensive. This is due to the fact that the approach demonstrated here will allow the sampling of those rare occurrences more frequently without misrepresenting their probabilistic impacts.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Likelihood-informed dimension reduction for nonlinear inverse problems

The intrinsic dimensionality of an inverse problem is affected by prior information, the accuracy and number of observations, and the smoothing properties of the forward operator. From a Bayesian perspective, changes from the prior to the posterior may, in many problems, be confined to a relatively low-dimensional subspace of the parameter space. We present a dimension reduction approach that defines and identifies such a subspace, called the 'likelihood-informed subspace' (LIS), by characterizing the relative influences of the prior and the likelihood over the support of the posterior distribution. This identification enables new and more efficient computational methods for Bayesian inference with nonlinear forward models and Gaussian priors. In particular, we approximate the posterior distribution as the product of a lower-dimensional posterior defined on the LIS and the prior distribution marginalized onto the complementary subspace. Markov chain Monte Carlo sampling can then proceed in lower dimensions, with significant gains in computational efficiency. We also introduce a Rao–Blackwellization strategy that de-randomizes Monte Carlo estimates of posterior expectations for additional variance reduction. We demonstrate the efficiency of our methods using two numerical examples: inference of permeability in a groundwater system governed by an elliptic PDE, and an atmospheric remote sensing problem based on Global Ozone Monitoring System (GOMOS) observations.

42 ENGINEERING↗

Physics-Informed Active Learning With Simultaneous Weak-Form Latent Space Dynamics Identification

The parametric greedy latent space dynamics identification (gLaSDI) framework has demonstrated promising potential for accurate and efficient modeling of high-dimensional nonlinear physical systems. However, it remains challenging to handle noisy data. Here, to enhance robustness against noise, we incorporate the weak-form estimation of nonlinear dynamics (WENDy) into gLaSDI. In the proposed weak-form gLaSDI (WgLaSDI) framework, an autoencoder and WENDy are trained simultaneously to discover intrinsic nonlinear latent-space dynamics of high-dimensional data. Compared with the standard sparse identification of nonlinear dynamics (SINDy) employed in gLaSDI, WENDy enables variance reduction and robust latent space discovery, therefore leading to more accurate and efficient reduced-order modeling. Furthermore, the greedy physics-informed active learning in WgLaSDI enables adaptive sampling of optimal training data on the fly for enhanced modeling accuracy. The effectiveness of the proposed framework is demonstrated by modeling various nonlinear dynamical problems, including viscous and inviscid Burgers' equations, time-dependent radial advection, and the Vlasov equation for plasma physics. With data that contains 5%–10% Gaussian white noise, WgLaSDI outperforms gLaSDI by orders of magnitude, achieving 1%–7% relative errors. Compared with the high-fidelity models, WgLaSDI achieves 121 to 1779x speed-up.

97 MATHEMATICS AND COMPUTING↗

Dispersion-enhanced sequential batch sampling for adaptive contour estimation

In computer simulation and optimal design, sequential batch sampling offers an appealing way to iteratively stipulate optimal sampling points based upon existing selections and efficiently construct surrogate modeling. Nonetheless, the issue of near duplicates poses tremendous quandary for sequential learning. It refers to the situation that selected critical points cluster together in each sampling batch, which are individually but not collectively informative towards the optimal design. Near duplicates severely diminish the computational efficiency as they barely contribute extra information towards update of the surrogate. To address this issue, we impose a dispersion criterion on concurrent selection of sampling points, which essentially forces a sparse distribution of critical points in each batch, and demonstrate the effectiveness of this approach in adaptive contour estimation. Specifically, we adopt Gaussian process surrogate to emulate the simulator, acquire variance reduction of the critical region from new sampling points as a dispersion criterion, and combine it with the modified expected improvement (EI) function for critical batch selection. The critical region here is the proximity of the contour of interest. This proposed approach is vindicated in numerical examples of a two-dimensional four-branch function, a four-dimensional function with a disjoint contour of interest and a time-delay dynamic system.

97 MATHEMATICS AND COMPUTING↗

Modeling rarefied gas chemistry with QuiPS, a novel quasi-particle method

The goal of this work is to build up the capability of quasi-particle simulation (QuiPS), a novel flow solver, such that it can adequately model the rarefied portion of an atmospheric reentry trajectory. Direct simulation Monte Carlo (DSMC) is the conventional solver for such conditions, but struggles to resolve transient flows, trace species, and high-level internal energy states due to stochastic noise. Quasi-particle simulation (QuiPS) is a novel Boltzmann solver that describes a system with a discretized, truncated velocity distribution function. The resulting fixed-velocity, variable weight quasi-particles enable smooth variation of macroscopic properties. The distribution function description enables the use of a variance-reduced collision model, greatly minimizing expense near equilibrium. This work presents the addition of a neutral air chemistry model to QuiPS and some demonstrative 0D simulations. The explicit representation of internal distributions in QuiPS reveals some of the flaws in existing physics models. Overall, variance reduction, a key feature of QuiPS, can greatly reduce expense of multi-dimensional calculations, but is only cheaper when the gas composition is near chemical equilibrium.

42 ENGINEERING↗

Dose Maps In/Around Target and Test Cell for DONES During the Operation

DONES (DEMO-Oriented Neutron Source) is to be constructed in order to investigate materials that are intended to be used for DEMO (DEMOnstration Power Plant) since they will be irradiated with very high fluxes of neutrons. A deuteron beam will bombard a lithium target situated in the so called Test Cell (TC) shielded with concrete. Monte Carlo simulations by means of the MCNP (Monte Carlo N-Particle) code were carried out to calculate both horizontal and vertical radiation dose maps in and around of the TC facility. Three geometrical configurations of the shielding and two kinds of concrete (ordinary and magnetite ones) were examined. For each case a spatial dose distribution and an accompanying errors map were provided. Advanced techniques of variance reduction were applied to compute doses in and outside of thick concrete walls. On the grounds of the obtained results a final configuration of the shielding was selected.

07 ISOTOPE AND RADIATION SOURCES↗

Weak-form latent space dynamics identification

Recent work in data-driven modeling has demonstrated that a weak formulation of model equations enhances the noise robustness of a wide range of computational methods. In this paper, we demonstrate the power of the weak form to enhance the LaSDI (Latent Space Dynamics Identification) algorithm, a recently developed data-driven reduced order modeling technique. We introduce a weak form-based version WLaSDI (Weak-form Latent Space Dynamics Identification). WLaSDI first compresses data, then projects onto the test functions and learns the local latent space models. Notably, WLaSDI demonstrates significantly enhanced robustness to noise. With WLaSDI, the local latent space is obtained using weak-form equation learning techniques. Compared to the standard sparse identification of nonlinear dynamics (SINDy) used in LaSDI, the variance reduction of the weak form guarantees a robust and precise latent space recovery, hence allowing for a fast, robust, and accurate simulation. We demonstrate the efficacy of WLaSDI vs. LaSDI on several common benchmark examples including viscid and inviscid Burgers', radial advection, and heat conduction. For instance, in the case of 1D inviscid Burgers' simulations with the addition of up to 100% Gaussian white noise, the relative error remains consistently below 6% for WLaSDI, while it can exceed 10,000% for LaSDI. Similarly, for radial advection simulations, the relative errors stay below 15% for WLaSDI, in stark contrast to the potential errors of up to 10,000% with LaSDI. Moreover, speedups of several orders of magnitude can be obtained with WLaSDI. For example applying WLaSDI to 1D Burgers' yields a 140X speedup compared to the corresponding full order model.

97 MATHEMATICS AND COMPUTING↗

Shutdown dose rate analysis with the Shift Monte Carlo radiation transport code and modular verification workflow

Calculation of the shutdown dose rate is crucial for safe fusion reactor operations. The Rigorous-two-step (R2S) method is a method that requires connected neutron transport, activation, and gamma transport. Shift has integrated variance reduction with a deterministic solver Denovo, supports multiple geometry formats, and is scalable. These features make it an attractive transport solver choice for an R2S workflow. An R2S workflow for the Shift Monte Carlo code is developed and compared to the existing Oak Ridge National Laboratory Shutdown Dose Rate Code Suite (ORCS) workflow. Also, a Python framework for integrating two R2S workflows is developed to mix and match each step in the R2S workflow for improved collaboration and verification experience. In this study, results show that the Shift-Denovo R2S workflow and the ORCS workflow calculate the shutdown dose rate of the ITER Shutdown Dose Rate benchmark problem with an average relative error of 2.285%.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An efficient, robust, domain-decomposition algorithm for particle Monte Carlo

A previously described algorithm [T.A. Brunner, T.J. Urbatsch, T.M. Evans, N.A. Gentile, Comparison of four parallel algorithms for domain decomposed implicit Monte Carlo, Journal of Computational Physics 212 (2) (2006) 527-539] for doing domain decomposed particle Monte Carlo calculations in the context of thermal radiation transport has been improved. It has been extended to support cases where the number of particles in a time step are unknown at the beginning of the time step. This situation arises when various physical processes, such as neutron transport, can generate additional particles during the time step, or when particle splitting is used for variance reduction. Additionally, several race conditions that existed in the previous algorithm and could cause code hangs have been fixed. This new algorithm is believed to be robust against all race conditions. The parallel scalability of the new algorithm remains excellent.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

An Extension of Implicit Monte Carlo Diffusion: Multigroup and The Difference Formulation

Implicit Monte Carlo (IMC) and Implicit Monte Carlo Diffusion (IMD) are approaches to the numerical solution of the equations of radiative transfer. IMD was previously derived and numerically tested on grey, or frequency-integrated problems. In this research, we extend Implicit Monte Carlo Diffusion (IMD) to account for frequency dependence, and we implement the difference formulation as a source manipulation variance reduction technique. We derive the relevant probability distributions and present the frequency dependent IMD algorithm, with and without the difference formulation. The IMD code with and without the difference formulation was tested using both grey and frequency dependent benchmark problems. The Su and Olson semi-analytic Marshak wave benchmark was used to demonstrate the validity of the code for grey problems. The Su and Olson semi-analytic picket fence benchmark was used for the frequency dependent problems. The frequency dependent IMD algorithm reproduces the results of both Su and Olson benchmark problems. Frequency group refinement studies indicate that the computational cost of refining the group structure is likely less than that of group refinement in deterministic solutions of the radiation diffusion methods. Our results show that applying the difference formulation to the IMD algorithm can result in an overall increase in the figure of merit for frequency dependent problems. However, the creation of negatively weighted particles from the difference formulation can cause significant numerical instabilities in regions of the problem with sharp spatial gradients in the solution. An adaptive implementation of the difference formulation may be necessary to focus its use in regions that are at or near thermal equilibrium.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Energy and momentum preserving Coulomb collision model for kinetic Monte Carlo simulations of plasma steady states in toroidal fusion devices

A kinetic Monte Carlo model suited for self-consistent transport studies is proposed and tested. The Monte Carlo collision operator is based on a widely used model of Coulomb scattering by a drifting Maxwellian and a new algorithm enforcing the momentum and energy conservation laws. The difference to other approaches consists in a specific procedure of calculating the background Maxwellian parameters, which does not require ensemble averaging and, therefore, allows for the use of single-particle algorithms. This possibility is useful in transport balance (steady state) problems with a phenomenological diffusive ansatz for the turbulent transport, because it allows a direct use of variance reduction methods well suited for single particle algorithms. In addition, a method for the self-consistent calculation of the electric field is discussed. Results of testing of the new collision operator using a set of 1D examples, and preliminary results of 2D modelling in realistic tokamak geometry, are presented.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Implementation, capabilities, and benchmarking of Shift, a massively parallel Monte Carlo radiation transport code

This paper discusses the implementation, capabilities, and validation of Shift, a massively parallel Monte Carlo radiation transport package developed and maintained at Oak Ridge National Laboratory. It has been developed to scale well from laptop to small computing clusters to advanced supercomputers. Special features of Shift include hybrid capabilities for variance reduction such as CADIS and FW-CADIS, and advanced parallel decomposition and tally methods optimized for scalability on supercomputing architectures. Shift has been validated and verified against various reactor physics benchmarks and compares well to other state-of-the-art Monte Carlo radiation transport codes such as MCNP5, CE KENO-VI, and OpenMC. Some specific benchmarks used for verification and validation include the CASL VERA criticality test suite and several Westinghouse AP1000 ® problems. These benchmark and scaling studies show promising results.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

A low-rank control variate for multilevel Monte Carlo simulation of high-dimensional uncertain systems

Multilevel Monte Carlo (MLMC) is a recently proposed variation of Monte Carlo (MC) simulation that achieves variance reduction by simulating the governing equations on a series of spatial (or temporal) grids with increasing resolution. Instead of directly employing the fine grid solutions, MLMC estimates the expectation of the quantity of interest from the coarsest grid solutions as well as differences between each two consecutive grid solutions. When the differences corresponding to finer grids become smaller, hence less variable, fewer MC realizations of finer grid solutions are needed to compute the difference expectations, thus leading to a reduction in the overall work. Here this paper presents an extension of MLMC, referred to as multilevel control variates (MLCV), where a low-rank approximation to the solution on each grid, obtained primarily based on coarser grid solutions, is used as a control variate for estimating the expectations involved in MLMC. Cost estimates as well as numerical examples are presented to demonstrate the advantage of this new MLCV approach over the standard MLMC when the solution of interest admits a low-rank approximation and the cost of simulating finer grids grows fast.

97 MATHEMATICS AND COMPUTING↗

Estimation of distributions via multilevel Monte Carlo with stratified sampling

We design and implement a novel algorithm for computing a multilevel Monte Carlo (MLMC) estimator of the joint cumulative distribution function (CDF) of a vector-valued quantity of interest in problems with random input parameters and initial conditions. Our approach combines MLMC with stratified sampling of the input sample space by replacing standard Monte Carlo at each level with stratified Monte Carlo initialized with proportionally allocated samples. We show that the resulting stratified MLMC (sMLMC) algorithm is more efficient than its standard MLMC counterpart due to the additional variance reduction provided by the stratification of the random parameter's domain, especially at the coarsest levels. Additional computational cost savings are obtained by smoothing the indicator function with a Gaussian kernel, which proves to be an efficient and robust alternative to recently developed polynomial-based techniques.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

UQpy: A general purpose Python package and development environment for uncertainty quantification

In this paper, we present the UQpy software toolbox, an open-source Python package for general uncertainty quantification (UQ) in mathematical and physical systems. The software serves as both a user-ready toolbox that includes many of the latest methods for UQ in computational modeling and a convenient development environment for Python programmers advancing the field of UQ. The paper presents an introduction to the software's architecture and existing capabilities, divided in the code in a set of modules centered around different UQ tasks such as sampling methods, generation of random processes and random fields, probabilistic inverse modeling, reliability analysis, surrogate modeling, and active learning. The paper also highlights the importance of the RunModel module, which is used to drive simulations in the uncertainty analyses performed in UQpy. This module conveniently allows the user to define computational models directly in Python, or to run simulations from a third-party software in serial or in parallel. To illustrate the various capabilities, two examples are tracked throughout the paper and analyzed repeatedly for various UQ tasks. The first is a Python model solving a nonlinear structural dynamics problem, used to illustrate UQpy's capabilities in sampling and forward propagation of high dimensional random vectors (stochastic processes), and probabilistic inference. The second model is a third-party Abaqus finite element model solving the thermomechanical response of a beam structure. This example is used to illustrate UQpy's capabilities in variance reduction sampling techniques, reliability analysis, surrogate modeling and active learning techniques.

97 MATHEMATICS AND COMPUTING↗

Electron Accelerator Shielding Design of KIPT Neutron Source Facility

The Argonne National Laboratory of the United States and the Kharkov Institute of Physics and Technology of the Ukraine have been collaborating on the design, development and construction of a neutron source facility at Kharkov Institute of Physics and Technology utilizing an electron-accelerator-driven subcritical assembly. The electron beam power is 100 kW using 100-MeV electrons. The facility was designed to perform basic and applied nuclear research, produce medical isotopes, and train nuclear specialists. The biological shield of the accelerator building was designed to reduce the biological dose to less than 5.0e-03 mSv/h during operation. The main source of the biological dose for the accelerator building is the photons and neutrons generated from different interactions of leaked electrons from the electron gun and the accelerator sections with the surrounding components and materials. The Monte Carlo N-particle extended code (MCNPX) was used for the shielding calculations because of its capability to perform electron-, photon-, and neutron-coupled transport simulations. The photon dose was tallied using the MCNPX calculation, starting with the leaked electrons. However, it is difficult to accurately tally the neutron dose directly from the leaked electrons. The neutron yield per electron from the interactions with the surrounding components is very small, similar to 0.01 neutron for 100-MeV electron and even smaller for lower-energy electrons. This causes difficulties for the Monte Carlo analyses and consumes tremendous computation resources for tallying the neutron dose outside the shield boundary with an acceptable accuracy. To avoid these difficulties, the SOURCE and TALLYX user subroutines of MCNPX were utilized for this study. The generated neutrons were banked, together with all related parameters, for a subsequent MCNPX calculation to obtain the neutron dose. The weight windows variance reduction technique was also utilized for both neutron and photon dose calculations. Two shielding materials, heavy concrete and ordinary concrete, were considered for the shield design. The main goal is to maintain the total dose outside the shield boundary less than 5.0e-03 mSv/h during operation. The shield configuration and parameters of the accelerator building were determined and are presented in this paper. Copyright (C) 2016, Published by Elsevier Korea LLC on behalf of Korean Nuclear Society.

43 PARTICLE ACCELERATORS↗

Photon detector response function methodology using MCNP and shift hybrid radiation transport code for wide-area contamination assay applications

Here, radiation transport modeling using the Monte Carlo N-Particle (MCNP) radiation transport code and Monte Carlo code, Shift, were employed to model detector responses for a variety of wide-area photon contamination scenarios. In this study, 2" × 2" and 3" × 3" cylindrical NaI(Tl) scintillation detector configurations at source detector-distances of 0.5 cm, 1 cm, 2.54 cm, 10 cm, and 30 cm were modeled. Media of soil, concrete, and steel were evaluated for contamination depths ranging from surface to a depth of an infinite thickness in each medium for photon energies ranging from 20 keV to 3 MeV, which correspond to the energies that current detectors can discern. Monoenergetic photon surface contamination detector responses for each of the media, source–detector distances, and detectors were estimated using MCNP v6.2. Shift was harnessed for improved variance reduction of particle transport in highly attenuating media to obtain average cell fluxes in the two MCNP NaI(Tl) scintillation detector configurations. Average cell flux values in Shift were coupled with detector responses from MCNP to convert average cell flux in a void to energy distribution of pulses in the NaI(Tl) scintillation detector crystal of interest. An optimized detector response function methodology was developed by coupling the MCNP radiation transport method with the Consistent Adjoint Driven Importance Sampling (CADIS) hybrid radiation transport method built into Shift to significantly decrease the runtime of thousands of MCNP pulse height simulations. The methodology may be utilized to quickly and accurately facilitate the assessment of a broad range of wide-area environmental contamination assay and decommissioning cleanup applications.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗