Search NASA⌕ Search

SEARCH · Search NASA

Results for “stochastic sampling”

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 127 records · Page 7

Derivative-free stochastic optimization via adaptive sampling strategies

In this paper, we present a novel derivative-free framework for solving unconstrained stochastic optimization problems. Many problems in fields ranging from simulation optimization to reinforcement learning to quantum computing involve settings where only stochastic function values are obtained via a zeroth-order oracle, which has no available gradient information and necessitates the usage of derivative-free optimization methodologies. Our approach includes estimating gradients using stochastic function evaluations and integrating adaptive sampling techniques to control the accuracy in these stochastic approximations. Our framework encapsulates several gradient estimation techniques, including standard finite-difference, Gaussian smoothing, sphere smoothing, randomized coordinate finite-difference, and randomized subspace finite-difference methods. We provide theoretical convergence guarantees for our framework and analyze the worst-case iteration and sample complexities associated with each gradient estimation method. Finally, we demonstrate the empirical performance of the methods on logistic regression and nonlinear least squares problems.

Adaptive sampling↗

Stochastic Stability of Nonlinear Sampled Data Systems with a Jump Linear Controller

This paper analyzes the stability of a sampled- data system consisting of a deterministic, nonlinear, time- invariant, continuous-time plant and a stochastic, discrete- time, jump linear controller. The jump linear controller mod- els, for example, computer systems and communication net- works that are subject to stochastic upsets or disruptions. This sampled-data model has been used in the analysis and design of fault-tolerant systems and computer-control systems with random communication delays without taking into account the inter-sample response. To analyze stability, appropriate topologies are introduced for the signal spaces of the sampled- data system. With these topologies, the ideal sampling and zero-order-hold operators are shown to be measurable maps. This paper shows that the known equivalence between the stability of a deterministic, linear sampled-data system and its associated discrete-time representation as well as between a nonlinear sampled-data system and a linearized representation holds even in a stochastic framework.

Gonzalez, Oscar R.↗

A preconditioning technique for Schur complement systems arising in stochastic optimization

Deterministic sample average approximations of stochastic programming problems with recourse are suitable for a scenario-based parallelization. In this paper the parallelization is obtained by using an interior-point method and a Schur complement mechanism for the interior-point linear systems. However, the direct linear solves involving the dense Schur complement matrix are expensive, and adversely affect the scalability of this approach. We address this issue by proposing a stochastic preconditioner for the Schur complement matrix and by using Krylov iterative methods for the solution of the dense linear systems. The stochastic preconditioner is built based on a subset of existing scenarios and can be assembled and factorized on a separate process before the computation of the Schur complement matrix finishes on the remaining processes. The expensive factorization of the Schur complement is removed from the parallel execution flow and the scaling of the optimization solver is considerably improved with this approach. The spectral analysis indicates an exponentially fast convergence in probability to 1 of the eigenvalues of the preconditioned matrix with the number of scenarios incorporated in the preconditioner. Numerical experiments performed on the relaxation of a unit commitment problem show good performance, in terms of both the accuracy of the solution and the execution time.

99 GENERAL AND MISCELLANEOUS↗

Analysis of Time Dependent Data and PRA

In the past decades, several numerical simulation codes have been employed to simulate accident dynamics (e.g., RELAP5-3D, MELCOR, MAAP). In order to evaluate the impact of uncertainties into accident dynamics, several stochastic methodologies have been coupled with these codes. These stochastic methods range from classical Monte-Carlo and Latin Hypercube sampling to stochastic polynomial methods. Similar approaches have been introduced into the risk and safety community where stochastic methods (such as RAVEN, ADAPT, MCDET, ADS) have been coupled with safety analysis codes in order to evaluate the safety impact of timing and sequencing of events. These approaches are usually called Dynamic PRA or simulation-based PRA methods. These uncertainties and safety methods usually generate a large number of simulation runs (database storage may be on the order of gigabytes or higher). The scope of this paper is to present a broad overview of methods and algorithms that can be used to analyze and extract information from large data sets containing time dependent data. In this context, 'extracting information' means constructing input-output correlations, finding commonalities, and identifying outliers. Some of the algorithms presented here have been developed or are under development within the RAVEN statistical framework. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Analytical Approach to Evaluate Uncertainty in Discharged Inventory Estimates at the Hanford Site - 19012

Many waste sites (>400) located within the Central Plateau portion of the Hanford Site received liquid discharges during the nuclear weapons material production mission (1944-1988). The Hanford Soil Inventory Model (SIM) is developed to quantitatively evaluate the discharged inventory for important radionuclides and chemicals at a given waste site to support fate and transport modeling of contaminants. An analytic approach was developed to estimate inventory and associated uncertainty. The approach was implemented for SIM and compared to a prior stochastic implementation applying a Monte-Carlo sampling approach. The analytic approach provides exact mean and variance results and the run times are much faster compared to the Monte-Carlo sampling approach. With the improved solution method using the analytical approach, it is feasible to run inventory updates and reduce associated uncertainty ranges. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Variance-Reduced Accelerated First-Order Methods: Central Limit Theorems and Confidence Statements

In this paper, we consider a strongly convex stochastic optimization problem and propose three classes of variable sample-size stochastic first-order methods: (i) the standard stochastic gradient descent method, (ii) its accelerated variant, and (iii) the stochastic heavy-ball method. In each scheme, the exact gradients are approximated by averaging across an increasing batch size of sampled gradients. We prove that when the sample size increases at a geometric rate, the generated estimates converge in mean to the optimal solution at an analogous geometric rate for schemes (i)–(iii). Based on this result, we provide central limit statements, whereby it is shown that the rescaled estimation errors converge in distribution to a normal distribution with the associated covariance matrix dependent on the Hessian matrix, the covariance of the gradient noise, and the step length. If the sample size increases at a polynomial rate, we show that the estimation errors decay at a corresponding polynomial rate and establish the associated central limit theorems (CLTs). Under certain conditions, we discuss how both the algorithms and the associated limit theorems may be extended to constrained and nonsmooth regimes. As a result, we provide an avenue to construct confidence regions for the optimal solution based on the established CLTs and test the theoretical findings on a stochastic parameter estimation problem.

Lei, Jinlong↗

Priors on red galaxy stochasticity from hybrid effective field theory

ABSTRACT We investigate the stochastic properties of typical red galaxy samples in a controlled numerical environment. We use halo occupation distribution (HOD) modelling to create mock realizations of three separate bright red galaxy samples consistent with data sets used for clustering and lensing analyses in modern galaxy surveys. Second-order Hybrid Effective Field Theory (HEFT) is used as a field-level forward model to describe the full statistical distribution of these tracer samples, and their stochastic power spectra are directly measured and compared to the Poisson shot-noise prediction. While all of the galaxy samples we consider are hosted within haloes with sub-Poisson stochasticity, we observe that the galaxy samples themselves possess stochasticities that range from sub-Poisson to super-Poisson, in agreement with predictions from the halo model. As an application of our methodology, we place priors on the expected degree of non-Poisson stochasticity in cosmological analyses using such samples. We expect these priors will be useful in reducing the complexity of the full parameter space for future analyses using second-order Lagrangian bias models. More generally, the techniques outlined here present the first application of HEFT methods to characterize models of the galaxy–halo connection at the field level, revealing new connections between once-disparate modelling frameworks.

79 ASTRONOMY AND ASTROPHYSICS↗

Importance Sampling with Analog Scenarios for Stochastic Economic Dispatch

Increasing penetrations of renewable energy sources such as wind into power grids motivates the investigation of new approaches to 5-minute economic dispatch computation. In this poster, we investigate the application of importance sampling in the two-stage stochastic economic dispatch problem while using historical data to characterize uncertainty. In our numerical experiments, we observe that importance sampling performs better than the standard Monte Carlo sampling approach in preventing loss-of-load. While first-stage costs are comparable for both sampling methods, the second-stage costs are significantly reduced when the importance sampling method is employed.

97 MATHEMATICS AND COMPUTING↗

Data Analysis Approaches for the Risk-Informed Safety Margins Characterization Toolkit

In the past decades, several numerical simulation codes have been employed to simulate accident dynamics (e.g., RELAP5-3D, RELAP-7, MELCOR, MAAP). In order to evaluate the impact of uncertainties into accident dynamics, several stochastic methodologies have been coupled with these codes. These stochastic methods range from classical Monte-Carlo and Latin Hypercube sampling to stochastic polynomial methods. Similar approaches have been introduced into the risk and safety community where stochastic methods (such as RAVEN, ADAPT, MCDET, ADS) have been coupled with safety analysis codes in order to evaluate the safety impact of timing and sequencing of events. These approaches are usually called Dynamic PRA or simulation-based PRA methods. These uncertainties and safety methods usually generate a large number of simulation runs (database storage may be on the order of gigabytes or higher). The scope of this paper is to present a broad overview of methods and algorithms that can be used to analyze and extract information from large data sets containing time dependent data. In this context, “extracting information” means constructing input-output correlations, finding commonalities, and identifying outliers. Some of the algorithms presented here have been developed or are under development within the RAVEN statistical framework.

97 MATHEMATICS AND COMPUTING↗

Design techniques - Stochastic controllers

Analytic techniques aid in the design of nearly optimal linear time-varying sampled-data stochastic controllers. The techniques also aid in the simplification and automation of program designing for control computers.

Widnall, W. S.↗

ANALYZING STAR CLUSTER POPULATIONS WITH STOCHASTIC MODELS: THE HUBBLE SPACE TELESCOPE/WIDE FIELD CAMERA 3 SAMPLE OF CLUSTERS IN M83

The majority of clusters in the universe have masses well below 10{sup 5} M{sub Sun }. Hence, their integrated fluxes and colors can be affected by the presence or absence of a few bright stars introduced by stochastic sampling of the stellar mass function. Specific methods are being developed to extend the analysis of cluster energy distributions into the low-mass regime. In this paper, we apply such a method to real observations of star clusters, in the nearby spiral galaxy M83. We reassess the ages and masses of a sample of 1242 clusters for which UBVIH{alpha} fluxes were obtained from observations with the Wide Field Camera 3 instrument on board the Hubble Space Telescope. Synthetic clusters with known properties are used to characterize the limitations of the method (valid range and resolution in age and mass, method artifacts). The ensemble of color predictions of the discrete cluster models are in good agreement with the distribution of observed colors. We emphasize the important role of the H{alpha} data in the assessment of the fraction of young objects, particularly in breaking the age-extinction degeneracy that hampers an analysis based on UBVI data only. We find the mass distribution of the cluster sample to follow a power law of index -2.1 {+-} 0.2, and the distribution of ages a power law of index -1.0 {+-} 0.2 for log (M/ M{sub Sun }) > 3.5, and ages between 10{sup 7} and 10{sup 9} yr. An extension of our main method, which makes full use of the probability distributions of age and mass obtained for the individual clusters of the sample, is explored. It produces similar power-law slopes and will deserve further investigation. Although the properties derived for individual clusters significantly differ from those obtained with traditional, non-stochastic models in about 30% of the objects, the first-order aspect of the age and mass distributions is similar to those obtained previously for this M83 sample in the range of overlap of the studies. We extend the power-law description to lower masses with better mass and age resolution and without most of the artifacts produced by the classical method.

79 ASTRONOMY AND ASTROPHYSICS↗

An adaptive sampling augmented Lagrangian method for stochastic optimization with deterministic constraints

The primary goal of this paper is to provide an efficient solution algorithm based on the augmented Lagrangian framework for optimization problems with a stochastic objective function and deterministic constraints. Our main contribution is combining the augmented Lagrangian framework with adaptive sampling, resulting in an efficient optimization methodology validated with practical examples. To achieve the presented efficiency, here we consider inexact solutions for the augmented Lagrangian subproblems, and through an adaptive sampling mechanism, we control the variance in the gradient estimates. Furthermore, we analyze the theoretical performance of the proposed scheme by showing equivalence to a gradient descent algorithm on a Moreau envelope function, and we prove sublinear convergence for convex objectives and linear convergence for strongly convex objectives with affine equality constraints. The worst-case sample complexity of the resulting algorithm, for an arbitrary choice of penalty parameter in the augmented Lagrangian function, is $\mathscr{O}$(ϵ -3-δ ) , where ϵ > 0 is the expected error of the solution and δ > 0 is a user-defined parameter. If the penalty parameter is chosen to be $\mathscr{O}$(ϵ -1 ), we demonstrate that the result can be improved to $\mathscr{O}$(ϵ -2 ) , which is competitive with the other methods employed in the literature. Moreover, if the objective function is strongly convex with affine equality constraints, we obtain $\mathscr{O}$(ϵ -1 log(1/ϵ)) complexity. Finally, we empirically verify the performance of our adaptive sampling augmented Lagrangian framework in machine learning optimization and engineering design problems, including topology optimization of a heat sink with environmental uncertainty.

97 MATHEMATICS AND COMPUTING↗

Assimilating partial observation to enhance feedback control of stochastic dynamical systems

Here, in this paper, we present a novel methodology to tackle feedback optimal control problems in scenarios where the exact state of the controlled process is unknown. It integrates data assimilation techniques and optimal control solvers to manage partial observation of the state process, a common occurrence in practical scenarios. Traditional stochastic optimal control methods assume full state observation, which is often not feasible in real-world fluid dynamics control problems. Our approach underscores the significance of utilizing observational data to inform control policy design. Specifically, we introduce a kernel learning backward stochastic differential equation (SDE) filter to enhance data assimilation efficiency and propose a sample-wise stochastic optimization method within the stochastic maximum principle framework. We demonstrate the efficacy and accuracy of our method in the control of advection-diffusion-reaction flow problem and the Dubins airplane maneuvering problem with model uncertainty.

data driven↗

A FRAMEWORK FOR INTERPRETING FAST RADIO TRANSIENTS SEARCH EXPERIMENTS: APPLICATION TO THE V-FASTR EXPERIMENT

We define a framework for determining constraints on the detection rate of fast transient events from a population of underlying sources, with a view to incorporate beam shape, frequency effects, scattering effects, and detection efficiency into the metric. We then demonstrate a method for combining independent data sets into a single event rate constraint diagram, using a probabilistic approach to the limits on parameter space. We apply this new framework to present the latest results from the V-FASTR experiment, a commensal fast transients search using the Very Long Baseline Array (VLBA). In the 20 cm band, V-FASTR now has the ability to probe the regions of parameter space of importance for the observed Lorimer and Keane fast radio transient candidates by combining the information from observations with differing bandwidths, and properly accounting for the source dispersion measure, VLBA antenna beam shape, experiment time sampling, and stochastic nature of events. We then apply the framework to combine the results of the V-FASTR and Allen Telescope Array Fly's Eye experiments, demonstrating their complementarity. Expectations for fast transients experiments for the SKA Phase I dish array are then computed, and the impact of large differential bandwidths is discussed.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Adaptive Sampling-Based Bi-Fidelity Stochastic Trust Region Method for Stochastic Derivative-Free Optimization

Bi-fidelity stochastic optimization has gained increasing attention as an efficient approach to reduce computational costs by leveraging a low-fidelity (LF) model to optimize an expensive high-fidelity (HF) objective. In this paper, we propose ASTRO-BFDF, an adaptive sampling trust-region method specifically designed for unconstrained bi-fidelity stochastic derivative-free optimization problems. In ASTRO-BFDF, the LF function serves two purposes: (i) to identify better iterates for the HF function when the optimization process indicates a high correlation between them and (ii) to reduce the variance of the HF function estimates using bi-fidelity Monte Carlo (BFMC). The algorithm dynamically determines sample sizes while adaptively choosing between crude Monte Carlo and BFMC to balance the trade-off between optimization and sampling errors. We prove that the iterates generated by ASTRO-BFDF converge to a first-order stationary point almost surely. Additionally, we demonstrate the effectiveness of the proposed algorithm through numerical experiments on synthetic benchmarks and simulation optimization problems involving discrete event systems.

97 MATHEMATICS AND COMPUTING↗

Error Convergence Characterization for Stochastic Transport Methods

There has been increased interest in reliably and efficiently solving engineering systems involving uncertainty in recent years. Characterization of the convergence of errors within the problem, including statistical and bias errors, increases confidence that chosen solution methods are appropriate for a particular problem and quantity of interest (QoI), that implementation errors have not been made, and can aid users or algorithms in solving problems within a tolerance efficiently. We demonstrate a method for characterizing error convergence in a simple stochastic transport problem. Though the approach is applicable to problems with more complicated physics, our simple problem enables analytic QoI solutions so that error convergence behavior and the identification of limiting errors can be presented with high precision. A one-dimensional, two-material, neutral-particle, and absorption-only transport problem involving uniformly distributed uncertain total cross sections is presented. Random sampling (RS) and stochastic collocation (SC) are used as stochastic solvers which sample from the uncertain space and Monte Carlo (MC) transport is employed to solve transport results on the resulting realizations. Probability density functions (PDFs) are created using either RS solutions or samples taken from a polynomial chaos expansion (PCE) constructed with SC solutions. The convergence of each solver is characterized with respect to the QoI against analytic solutions, convergence behavior is used to efficiently solve a QoI within a tolerance, and convergence plots are used to identify dominant solution errors.

42 ENGINEERING↗