Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “stochastic methods”

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 1,405 records · Page 78
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

A biorthogonal decomposition for the identification and simulation of non-stationary and non-Gaussian random fields

In this paper, a new method for the identification and simulation of non-Gaussian and non-stationary stochastic fields given a database is proposed. It is based on two successive biorthogonal decompositions aiming at representing spatio–temporal stochastic fields. The proposed double expansion allows to build the model even in the case of large-size problems by separating the time, space and random parts of the field. A Gaussian kernel estimator is used to simulate the high dimensional set of random variables appearing in the decomposition. The capability of the method to reproduce the non-stationary and non-Gaussian features of random phenomena is illustrated by applications to earthquakes (seismic ground motion) and sea states (wave heights).

58 GEOSCIENCES↗
DOE OSTI2013

A Comparison of Forecast Error Generators for Modeling Wind and Load Uncertainty

This paper presents four algorithms to generate random forecast error time series, including a truncated-normal distribution model, a state-space based Markov model, a seasonal autoregressive moving average (ARMA) model, and a stochastic-optimization based model. The error time series are used to create real-time (RT), hour-ahead (HA), and day-ahead (DA) wind and load forecast time series that statistically match historically observed forecasting data sets, used for variable generation integration studies. A comparison is made using historical DA load forecast and actual load values to generate new sets of DA forecasts with similar stoical forecast error characteristics. This paper discusses and compares the capabilities of each algorithm to preserve the characteristics of the historical forecast data sets.

Lu, Ning↗
DOE OSTI2019

Advising Big Data Transfer Over Dedicated Connections Based on Profiling Optimization

Big data transfer in next-generation scientific applications is currently commonly carried out over dedicated channels in high-performance networks (HPNs), where transport protocols play a critical role in maximizing application-level throughput. Optimizing the performance of these protocols is challenging: i) transport protocols perform differently in various network environments, and the protocol choice is not straightforward; ii) even for a given protocol in a given environment, different parameter settings of the protocol may lead to significantly different performance and oftentimes the default setting does not yield the best performance. However, it is prohibitively time-consuming to conduct exhaustive transport profiling due to the large parameter space. In this paper, we propose a PRofiling Optimization Based DAta Transfer Advisor (ProbData) to help end users determine the most effective transport method with the most appropriate parameter settings to achieve satisfactory performance for big data transfer over dedicated connections in HPNs. ProbData employs a fast profiling scheme based on the Simultaneous Perturbation Stochastic Approximation algorithm, namely, FastProf, to accelerate the exploration of the optimal operational zones of various transport methods to improve profiling efficiency. We first present a theoretical background of the optimized profiling approach in ProbData and then detail its design and implementation. The advising procedure and performance benefits of FastProf and ProbData are illustrated and evaluated by both extensive emulations based on real-life performance measurements and experiments over various physical connections in existing production HPNs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

A Langevin approach to multi-scale modeling

In plasmas, distribution functions often demonstrate long anisotropic tails or otherwise significant deviations from local Maxwellians. The tails, especially if they are pulled out from the bulk, pose a serious challenge for numerical simulations as resolving both the bulk and the tail on the same mesh is often challenging. A multi-scale approach, providing evolution equations for the bulk and the tail individually, could offer a resolution in the sense that both populations could be treated on separate meshes or different reduction techniques applied to the bulk and the tail population. In this paper, we propose a multi-scale method which allows us to split a distribution function into a bulk and a tail so that both populations remain genuine, non-negative distribution functions and may carry density, momentum, and energy. The proposed method is based on the observation that the motion of an individual test particle in a plasma obeys a stochastic differential equation, also referred to as a Langevin equation. Finally, this allows us to define transition probabilities between the bulk and the tail and to provide evolution equations for both populations separately.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2013

A stochastic model of cell survival for high-Z nanoparticle radiotherapy

Purpose: The authors present a stochastic framework for the assessment of cell survival in gold nanoparticle radiotherapy. Methods: The authors derive the equations for the effective macroscopic dose enhancement for a population of cells with nonideal distribution of gold nanoparticles (GNP), allowing different number of GNP per cell and different distances with respect to the cellular target. They use the mixed Poisson distribution formalism to model the impact of the aforementioned physical factors on the effective dose enhancement. Results: The authors show relatively large differences in the estimation of cell survival arising from using approximated formulae. They predict degeneration of the cell killing capacity due to different number of GNP per cell and different distances with respect to the cellular target. Conclusions: The presented stochastic framework can be used in interpretation of experimental cell survival or tumor control probability studies.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Microgrid energy scheduling under uncertain extreme weather: Adaptation from parallelized reinforcement learning agents

Microgrids are useful solutions for integrating renewable energy resources and providing seamless green electricity to minimize carbon footprint. In recent years, extreme weather events happened often worldwide and caused significant economic and societal losses. Such events bring uncertainties to the microgrid energy scheduling problems and increase the challenges of microgrid operation. Traditional optimization approaches suffer from the inaccuracy of the uncertain microgrid model and the unseen events. Existing reinforcement learning (RL) - based approaches are also hampered by the limited generalization and the increasing computational burden when stochastic formulations are required to accommodate the uncertainties. This paper proposes a new parallelized reinforcement learning (PRL) method based on the probabilistic events to handle the microgrid energy uncertainties. Specifically, several local learning agents are employed to interact with pertinent microgrid environments in a distributed manner and report outcomes to the global agent, which will optimize microgrid energy resources online during extreme events. The stochastic microgrid energy optimization problem is reformulated to include all possible scenarios with probabilities. The advantage estimate functions of learning agents are designed with a backward sweep to transfer the outcomes to the value function updating process. Two simulation studies, stochastic optimization and online testing, are performed to compare with several existing RL approaches. Results substantiate that the proposed PRL method can achieve up to 20% optimization performance improvement with 4 and 28 times less computation cost than Q-learning with experience replay and multi-agent Q-learning approaches, respectively.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2019

Pilot Point Optimization of Mining Boundaries for Lateritic Metal Deposits: Finding the Trade-off Between Dilution and Ore Loss

Geological contacts in lateritic metal deposits (footwall topographies) often delineate the orebody boundaries. Spatial variations seen in such contacts are frequently higher than those for the metal grades of the deposit. Therefore, borehole spacing chosen based on the grade variations cannot adequately capture the geological contact variability. Consequently, models created using such boreholes cause high volumetric uncertainties in the actual and targeted ore extraction volumes, which, in turn, lead to high unplanned dilution and ore losses. In this paper, a method to design optimum ore/mining boundaries for lateritic metal deposits is presented. The proposed approach minimizes the dilution/ore losses and comprises two main steps. First, the uncertainty on the orebody boundary is represented using a set of stochastic realizations generated with a multiple-point statistics algorithm. Then, the optimal orebody boundary is determined using an optimization technique inspired by a model calibration method called Pilot Points. The pilot points represent synthetic elevation values, and they are used to construct smooth mining boundaries using the multilevel B-spline technique. The performance of a generated surface is evaluated using the expected sum of losses in each of the stochastic realizations. The simulated annealing algorithm is used to iteratively determine the pilot point values which minimize the expected losses. The results show a significant reduction in the dilution volume as compared to those obtained from the actual mining operation.

58 GEOSCIENCES↗
DOE OSTI2013

A diffusive information preservation method for small Knudsen number flows

The direct simulation Monte Carlo (DSMC) method is a powerful particle-based method for modeling gas flows. It works well for relatively large Knudsen (Kn) numbers, typically larger than 0.01, but quickly becomes computationally intensive as Kn decreases due to its time step and cell size limitations. An alternative approach was proposed to relax or remove these limitations, based on replacing pairwise collisions with a stochastic model corresponding to the Fokker–Planck equation [J. Comput. Phys., 229, 1077 (2010); J. Fluid Mech., 680, 574 (2011)]. Similar to the DSMC method, the downside of that approach suffers from computationally statistical noise. To solve the problem, a diffusion-based information preservation (D-IP) method has been developed. The main idea is to track the motion of a simulated molecule from the diffusive standpoint, and obtain the flow velocity and temperature through sampling and averaging the IP quantities. To validate the idea and the corresponding model, several benchmark problems with Kn ∼ 10{sup −3}–10{sup −4} have been investigated. It is shown that the IP calculations are not only accurate, but also efficient because they make possible using a time step and cell size over an order of magnitude larger than the mean collision time and mean free path, respectively.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Numerical methods for fractional Fokker–Planck equation with multiplicative Marcus Lévy noises

The Fokker–Planck equation (FPE) is an important deterministic tool for investigating stochastic dynamical systems. In this paper, we consider the space-time fractional FPE driven by multiplicative Marcus Lévy noises. Efficient numerical schemes are presented to solve the equations. Stability and convergence of the methods are also discussed. We give some numerical experiments to validate our schemes, and examine the effects of parameters on solutions. Additionally, we analyze the maximal likely trajectories and the critical time for the change of the most probability location.

Mathematics↗
NASA NTRS1997

Algorithms for Performance, Dependability, and Performability Evaluation using Stochastic Activity Networks

Modeling tools and technologies are important for aerospace development. At the University of Illinois, we have worked on advancing the state of the art in modeling by Markov reward models in two important areas: reducing the memory necessary to numerically solve systems represented as stochastic activity networks and other stochastic Petri net extensions while still obtaining solutions in a reasonable amount of time, and finding numerically stable and memory-efficient methods to solve for the reward accumulated during a finite mission time. A long standing problem when modeling with high level formalisms such as stochastic activity networks is the so-called state space explosion, where the number of states increases exponentially with size of the high level model. Thus, the corresponding Markov model becomes prohibitively large and solution is constrained by the the size of primary memory. To reduce the memory necessary to numerically solve complex systems, we propose new methods that can tolerate such large state spaces that do not require any special structure in the model (as many other techniques do). First, we develop methods that generate row and columns of the state transition-rate-matrix on-the-fly, eliminating the need to explicitly store the matrix at all. Next, we introduce a new iterative solution method, called modified adaptive Gauss-Seidel, that exhibits locality in its use of data from the state transition-rate-matrix, permitting us to cache portions of the matrix and hence reduce the solution time. Finally, we develop a new memory and computationally efficient technique for Gauss-Seidel based solvers that avoids the need for generating rows of A in order to solve Ax = b. This is a significant performance improvement for on-the-fly methods as well as other recent solution techniques based on Kronecker operators. Taken together, these new results show that one can solve very large models without any special structure.

Deavours, Daniel D.↗
NASA NTRS2021

Multi-Pass Sequential Mini-Batch Stochastic Gradient Descent Algorithms for Noise Covariance Estimation in Adaptive Kalman Filtering

Estimation of unknown noise covariances in a Kalman filter is a problem of significant practical interest in a wide array of applications. Although this problem has a long history, reliable algorithms for their estimation were scant, and necessary and sufficient conditions for identifiability of the covariances were in dispute until recently. Necessary and sufficient conditions for covariance estimation and a batch estimation algorithm were presented in our previous study. This paper presents stochastic gradient descent algorithms for noise covariance estimation in adaptive Kalman filters that are an order of magnitude faster than the batch method for similar or better root mean square error. More significantly, these algorithms are applicable to non-stationary systems where the noise covariances can occasionally jump up or down by an unknown magnitude. The computational efficiency of the new algorithms stems from adaptive thresholds for convergence, recursive fading memory estimation of the sample cross-correlations of the innovations, and accelerated stochastic gradient descent algorithms. The comparative evaluation of the proposed methods on a number of test cases demonstrates their computational efficiency and accuracy.

Adaptive Kalman filtering↗
DOE OSTI2019

Molten salt reactor core simulation with PROTEUS

The Nuclear Energy Advance Modeling and Simulation Program (NEAMS) of the US Department of Energy (DOE) aims to develop advanced simulation tools to aid the development of advanced nuclear technologies. The high-fidelity PROTEUS code was developed in the NEAMS program for reactor physics analysis. This paper tests the applicability of the PROTEUS code to simulate a Molten Salt Reactor (MSR). The OpenMC code based on the Monte Carlo Method was employed to generate the cross sections and to provide stochastic solutions for verification. The PROTEUS code was successfully employed to simulate the steady-state performance of an MSR represented by the ThorCon core design. The 2D lattice, 2D core, and 3D full core calculations performed by PROTEUS show good agreement with that predicted by OpenMC. Furthermore, for the nodal option within PROTEUS, the precursor drift model was also tested, and this effect resulted in a reduction of the k-eff by 300 pcm.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2025

Jensen–Shannon divergence based novel loss functions for Bayesian neural networks

Bayesian neural networks (BNNs) are state-of-the-art machine learning methods that can naturally regularize and systematically quantify uncertainties using their stochastic parameters. Kullback–Leibler (KL) divergence-based variational inference used in BNNs suffer from unstable optimization and challenges in approximating light-tailed posteriors due to the unbounded nature of the KL divergence. To resolve these issues, we formulate a novel loss function for BNNs based on a new modification to the generalized Jensen–Shannon (JS) divergence, which is bounded. In addition, we propose a Geometric JS divergence-based loss, which is computationally efficient since it can be evaluated analytically. We found that the JS divergence-based variational inference is intractable, and hence employed a constrained optimization framework to formulate these losses. Our theoretical analysis and empirical experiments on multiple regression and classification data sets suggest that the proposed losses perform better than the KL divergence-based loss, especially when the data sets are noisy or biased. Specifically, there are approximately 5% and 8% improvements in accuracy for a noise-added CIFAR-10 dataset and a regression dataset, respectively. There is about 13% reduction in false negative predictions of a biased histopathology dataset. Additionally, we quantify and compare the uncertainty metrics for the regression and classification tasks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2012

Transcriptional burst frequency and burst size are equally modulated across the human genome

Gene expression occurs either as an episodic process, characterized by pulsatile bursts or as a constitutive, Poisson-like accumulation of gene products. It is not clear which mode of gene expression (constitutive versus bursty) predominates across a genome or how transcriptional dynamics are influenced by genomic position and promoter sequence. Here, we use time-lapse fluorescence microscopy, building off of theoretical studies that exploit the time-resolved structure of stochastic fluctuations in gene expression, to develop a three-dimensional method for mapping underlying gene-regulatory mechanisms. Over 8,000 individual human genomic loci were analyzed, and at virtually all loci, episodic bursting as opposed to constitutive expression was found to be the predominant mode of expression. Quantitative analysis of the expression dynamics at these 8,000 loci indicates that both frequency and size of transcriptional bursts vary equally across the human genome independent of promoter sequence. Strikingly, weaker expression loci modulate burst frequency to increase activity, while stronger expression loci modulate burst size to increase activity. Transcriptional activators, such as TNF, generate similar patterns of change in burst frequency and burst size. In summary, transcriptional bursting dominates across the human genome, both burst frequency and burst size vary by chromosomal location, and transcriptional activators alter burst frequency and burst size, depending on the expression level of the locus.

noise biology↗
DOE OSTI2024

Explicit simulation of the Brownian rotation of arbitrary shaped aerosol particles using quaternions

The shape of an aerosol particle strongly influences its mass and momentum transfer cross-sections, charging properties, and other physical properties. Here, we present an explicit time-stepping procedure to simulate the rotational Brownian motion of arbitrary shaped aerosol particles by solving Euler’s equation of rotation. A Langevin formulation of the rotation equations is used, wherein Brownian motion due to thermal collisions between a particle and background gas molecules is represented using a stochastic fluctuating torque and fluid resistance is included as a drag torque. To avoid singularities associated with describing the orientation of a shape with Euler angles, we employ a quaternion formulation that leads to first-order stochastic differential equations to describe the evolution of the angular position and angular velocity of a rigid body. We perform all the rotational dynamics calculations in the body-fixed frame of reference attached to the rotating shape whose basis vectors are the normalized eigenvectors of the inertia tensor of the particle. Numerical solutions to rotation under torque-free conditions, damped rotation without Brownian motion, and stochastic rotation for arbitrary shapes are presented and discussed. The presented method enables time-resolved simulation of Brownian rotation for direct comparison with experimentally measured trajectories or statistical measures. The second order accuracy of the used time-stepping procedure places a severe restriction on the timestep that can be used for obtaining accurate results. Animations of presented simulations are included for visualizing rotational motion at various gas pressures. To aid implementation, MATLAB ® codes are also provided. Extension to include translation Brownian motion is straightforward.

Roy, Mrittika↗
DOE OSTI2020

SSA Monte Carlo and Master Equation Modeling of Neutron Leakage Distributions

Strongly stochastic neutron populations in highly subcritical multiplying media containing weak spontaneous fission sources are unique indicators of the presence of SNM. For instance, in neutron multiplicity counting, an extended tail in the count number probability distribution is observed that is a characteristic of and sensitive to the amount of SNM present. Theoretical modeling and efficient computation of stochastic neutron distributions are complementary to experimental measurements and moreover shed light on the fundamental interactions and transport processes that underlie nuclear signatures. Modeling approaches that are intermediate between gross-system level models and high-fidelity but computationally demanding Monte Carlo simulations are of particular interest in this regard. Furthermore we present three advances on existing work in point kinetic or lumped geometry: (i) implementation of the Stochastic Simulation Algorithm (SSA), a system state-updating Monte Carlo method that is more e cient than single-event Monte Carlo in lumped geometries, (ii) application of the backward Master equation to obtain time-gated count moments, and (iii) reconstructing the count number distribution from the deterministically computed moments using a generalized Laguerre polynomial expansion.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1988

Some comments on particle image displacement velocimetry

Laser speckle velocimetry (LSV) or particle image displacement velocimetry, is introduced. This technique provides the simultaneous visualization of the two-dimensional streamline pattern in unsteady flows as well as the quantification of the velocity field over an entire plane. The advantage of this technique is that the velocity field can be measured over an entire plane of the flow field simultaneously, with accuracy and spatial resolution. From this the instantaneous vorticity field can be easily obtained. This constitutes a great asset for the study of a variety of flows that evolve stochastically in both space and time. The basic concept of LSV; methods of data acquisition and reduction, examples of its use, and parameters that affect its utilization are described.

Lourenco, L. M.↗
NASA NTRS1990

Turbulence modulation effect on evaporating spray characterization

The effects of turbulence modulation on the evaporating spray characteristics of a liquid fuel jet are numerically investigated. Two modulation models are implemented into the k-epsilon two-equation model. The gas phase is solved by a pressure-based method on a Eulerian frame and droplets are solved by a stochastic discrete particle technique based on Lagrangian formulation. The k-epsilon models are used to characterize the time and length scales of the gas phase turbulence for droplet dispersions and droplet/turbulence interactions. The modulation models are found to have significant effects on gas phase dynamics, and are less important for the droplet dynamics. Numerical predictions are compared with the experimental data of Drallemier and Peters (1990).

Shang, H. M.↗