Search NASA⌕ Search

SEARCH · Search NASA

Results for “stochastic approximation”

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 991 records · Page 55

Employing Sensitivity Derivatives for Robust Optimization under Uncertainty in CFD

A robust optimization is demonstrated on a two-dimensional inviscid airfoil problem in subsonic flow. Given uncertainties in statistically independent, random, normally distributed flow parameters (input variables), an approximate first-order statistical moment method is employed to represent the Computational Fluid Dynamics (CFD) code outputs as expected values with variances. These output quantities are used to form the objective function and constraints. The constraints are cast in probabilistic terms; that is, the probability that a constraint is satisfied is greater than or equal to some desired target probability. Gradient-based robust optimization of this stochastic problem is accomplished through use of both first and second-order sensitivity derivatives. For each robust optimization, the effect of increasing both input standard deviations and target probability of constraint satisfaction are demonstrated. This method provides a means for incorporating uncertainty when considering small deviations from input mean values.

Newman, Perry A.↗

Self‐Potential Tomography Preconditioned by Particle Swarm Optimization—Application to Monitoring Hyporheic Exchange in a Bedrock River

Abstract A self‐potential (SP) data‐inversion algorithm was developed and tested on an analytical model of electrical‐potential profile data attributed to single and multiple polarized electrical sources. The developed algorithm was then validated by an application to SP‐monitoring field data measured on the floodplain of East Fork Poplar Creek, Oak Ridge, Tennessee, to image electrical sources in areas conducive to preferential flow into the flood plain from the bedrock‐lined riverbed. The algorithm combined stochastic source‐localization by particle‐swarm‐optimization (PSO) of electrical sources characterized by simplified geometries with source tomography by regularized weighted least‐squares minimization of a quadratic objective function. Prior information was incorporated by preconditioning the tomography algorithm by PSO results. Variable percentages of random noise were added to analytical‐model data to evaluate the algorithm performance. Results indicated that true parameters of single‐source models were inverted and approximated with small residual error, whereas inversion of analytical‐model data representing multiple electrical sources accurately approximated the locations of the sources but miscalculated some parameters because of the non‐uniqueness of the inverse‐model solution. Source tomography applied to analytical model data during testing produced a spatially continuous parameter field that identified the locations of point‐scale synthetic dipole sources of electrical current flow with varying degrees of accuracy depending on the prior information incorporated into the tomography. When applied to SP‐monitoring field data, the algorithm imaged electrical sources within a known fault that intersects the bedrock riverbed and flood plain of East Fork Poplar Creek and depicted dynamic electrical conditions attributed to hyporheic exchange.

54 ENVIRONMENTAL SCIENCES↗

Green's function-stochastic methods framework for probing nonlinear evolution problems: Burger's equation, the nonlinear Schroedinger's equation, and hydrodynamic organization of near-molecular-scale vorticity

Research Highlights: > Systematic approach for physically probing nonlinear and random evolution problems. > Evolution of vortex sheets corresponds to evolution of an Ornstein-Uhlenbeck process. > Organization of near-molecular scale vorticity mediated by hydrodynamic modes. > Framework allows calculation of vorticity evolution within random strain fields. - Abstract: A framework which combines Green's function (GF) methods and techniques from the theory of stochastic processes is proposed for tackling nonlinear evolution problems. The framework, established by a series of easy-to-derive equivalences between Green's function and stochastic representative solutions of linear drift-diffusion problems, provides a flexible structure within which nonlinear evolution problems can be analyzed and physically probed. As a preliminary test bed, two canonical, nonlinear evolution problems - Burgers' equation and the nonlinear Schroedinger's equation - are first treated. In the first case, the framework provides a rigorous, probabilistic derivation of the well known Cole-Hopf ansatz. Likewise, in the second, the machinery allows systematic recovery of a known soliton solution. The framework is then applied to a fairly extensive exploration of physical features underlying evolution of randomly stretched and advected Burger's vortex sheets. Here, the governing vorticity equation corresponds to the Fokker-Planck equation of an Ornstein-Uhlenbeck process, a correspondence that motivates an investigation of sub-sheet vorticity evolution and organization. Under the assumption that weak hydrodynamic fluctuations organize disordered, near-molecular-scale, sub-sheet vorticity, it is shown that these modes consist of two weakly damped counter-propagating cross-sheet acoustic modes, a diffusive cross-sheet shear mode, and a diffusive cross-sheet entropy mode. Once a consistent picture of in-sheet vorticity evolution is established, a number of analytical results, describing the motion and spread of single, multiple, and continuous sets of Burger's vortex sheets, evolving within deterministic and random strain rate fields, under both viscous and inviscid conditions, are obtained. In order to promote application to other nonlinear problems, a tutorial development of the framework is presented. Likewise, time-incremental solution approaches and construction of approximate, though otherwise difficult-to-obtain backward-time GF's (useful in solution of forward-time evolution problems) are discussed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulating the universe(s): from cosmic bubble collisions to cosmological observables with numerical relativity

The theory of eternal inflation in an inflaton potential with multiple vacua predicts that our universe is one of many bubble universes nucleating and growing inside an ever-expanding false vacuum. The collision of our bubble with another could provide an important observational signature to test this scenario. We develop and implement an algorithm for accurately computing the cosmological observables arising from bubble collisions directly from the Lagrangian of a single scalar field. We first simulate the collision spacetime by solving Einstein's equations, starting from nucleation and ending at reheating. Taking advantage of the collision's hyperbolic symmetry, the simulations are performed with a 1+1-dimensional fully relativistic code that uses adaptive mesh refinement. We then calculate the comoving curvature perturbation in an open Friedmann-Robertson-Walker universe, which is used to determine the temperature anisotropies of the cosmic microwave background radiation. For a fiducial Lagrangian, the anisotropies are well described by a power law in the cosine of the angular distance from the center of the collision signature. For a given form of the Lagrangian, the resulting observational predictions are inherently statistical due to stochastic elements of the bubble nucleation process. Further uncertainties arise due to our imperfect knowledge about inflationary and pre-recombination physics. We characterize observational predictions by computing the probability distributions over four phenomenological parameters which capture these intrinsic and model uncertainties. This represents the first fully-relativistic set of predictions from an ensemble of scalar field models giving rise to eternal inflation, yielding significant differences from previous non-relativistic approximations. Thus, our results provide a basis for a rigorous confrontation of these theories with cosmological data.

79 ASTRONOMY AND ASTROPHYSICS↗

Stochastic Guidance of Buoyancy Controlled Vehicles under Ice Shelves using Ocean Currents

We propose a novel technique for guidance ofbuoyancy-controlled vehicles in uncertain under-ice ocean flows.In-situ melt rate measurements collected at the grounding zoneof Antarctic ice shelves, where the ice shelf meets the underlyingbedrock, are essential to constrain models of future sea levelrise. Buoyancy-controlled vehicles, which control their verticalposition in the water column but have no means of horizontalpropulsion, offer an affordable and reliable platform for suchin-situ data collection. However, reaching the grounding zonerequires vehicles to traverse tens of kilometers under the iceshelf, with approximate position knowledge and no meansof communication, in highly variable and uncertain oceancurrents. To address this challenge, we propose a partiallyobservable MDP approach that exploits model-based knowledgeof the under-ice currents and, critically, of their uncertainty,to synthesize effective guidance policies. The approach usesapproximate dynamic programming to model uncertainty inthe currents, and QMDP to address localization uncertainty.Numerical experiments show that the policy can deliver upto 88.8% of underwater vehicles to the grounding zone – a33% improvement compared to state-of-the-art guidance techniques, and a 262% improvement over uncontrolled drifters.Collectively, these results show that model-based under-iceguidance is a highly promising technique for exploration ofunder-ice cavities, and has the potential to enable cost-effectiveand scalable access to these challenging and rarely observed environments.

Clark, Evan B.↗

Classical shadows for quantum process tomography on near-term quantum computers

Quantum process tomography is a powerful tool for understanding quantum channels and characterizing the properties of quantum devices. Inspired by recent advances using classical shadows in quantum state tomography [H.-Y. Huang, R. Kueng, and J. Preskill, .], we have developed ShadowQPT, a classical shadow method for quantum process tomography. We introduce two related formulations with and without ancilla qubits. ShadowQPT stochastically reconstructs the Choi matrix of the device allowing for an classical evaluation of the device on arbitrary inputs with respect to arbitrary outputs. Using shadows, we then show how to compute overlaps, generate all k -weight reduced processes, and perform reconstruction via Hamiltonian learning. These latter two tasks are efficient for large systems as the number of quantum measurements needed scales only logarithmically with the number of qubits. A number of additional approximations and improvements are developed, including the use of a pair-factorized Clifford shadow and a series of postprocessing techniques that significantly enhance the accuracy for recovering the quantum channel. We have implemented ShadowQPT using both Pauli and Clifford measurements on the IonQ trapped ion quantum computer for quantum processes up to n = 4 qubits, and we achieved good performance. Published by the American Physical Society 2024

Levy, Ryan (ORCID:0000000349527156)↗

An efficient Bayesian data-worth analysis using a multilevel Monte Carlo method

Improving the understanding of subsurface systems and thus reducing prediction uncertainty requires collection of data. As the collection of subsurface data is costly, it is important that the data collection scheme is cost-effective. Design of a cost-effective data collection scheme, i.e., data-worth analysis, requires quantifying model parameter, prediction, and both current and potential data uncertainties. Assessment of these uncertainties in large-scale stochastic subsurface hydrological model simulations using standard Monte Carlo (MC) sampling or surrogate modeling is extremely computationally intensive, sometimes even infeasible. In this paper, we propose an efficient Bayesian data-worth analysis using a multilevel Monte Carlo (MLMC) method. Compared to the standard MC that requires a significantly large number of high-fidelity model executions to achieve a prescribed accuracy in estimating expectations, the MLMC can substantially reduce computational costs using multifidelity approximations. Since the Bayesian data-worth analysis involves a great deal of expectation estimation, the cost saving of the MLMC in the assessment can be outstanding. While the proposed MLMC-based data-worth analysis is broadly applicable, we use it for a highly heterogeneous two-phase subsurface flow simulation to select an optimal candidate data set that gives the largest uncertainty reduction in predicting mass flow rates at four production wells. Finally, the choices made by the MLMC estimation are validated by the actual measurements of the potential data, and consistent with the standard MC estimation. But compared to the standard MC, the MLMC greatly reduces the computational costs.

97 MATHEMATICS AND COMPUTING↗

Perspective: Explicitly correlated electronic structure theory for complex systems

The explicitly correlated approach is one of the most important breakthroughs in ab initio electronic structure theory, providing arguably the most compact, accurate, and efficient ansatz for describing the correlated motion of electrons. Since Hylleraas first used an explicitly correlated wave function for the He atom in 1929, numerous attempts have been made to tackle the significant challenges involved in constructing practical explicitly correlated methods that are applicable to larger systems. These include identifying suitable mathematical forms of a correlated wave function and an efficient evaluation of many-electron integrals. R12 theory, which employs the resolution of the identity approximation, emerged in 1985, followed by the introduction of novel correlation factors and wave function ansätze, leading to the establishment of F12 theory in the 2000s. Rapid progress in recent years has significantly extended the application range of explicitly correlated theory, offering the potential of an accurate wave-function treatment of complex systems such as photosystems and semiconductors. In conclusion, this perspective surveys explicitly correlated electronic structure theory, with an emphasis on recent stochastic and deterministic approaches that hold significant promise for applications to large and complex systems including solids.

74 ATOMIC AND MOLECULAR PHYSICS↗

Sliced-Inverse-Regression--Aided Rotated Compressive Sensing Method for Uncertainty Quantification

Compressive-sensing-based uncertainty quantification methods have become a pow- erful tool for problems with limited data. In this paper, we use the sliced inverse regression (SIR) method to provide an initial guess for the alternating direction method, which is used to enhance sparsity of the Hermite polynomial expansion of stochastic state variables. The sparsity improvement increases both the efficiency and accuracy of the compressive-sensing-based uncertainty quantification method. We demonstrate that the initial guess from SIR is more suitable for the cases when the available data is very limited (Algorithm 4). We also propose another algorithm (Algorithm 5) that performs dimension reduction first with sliced inverse regression, then constructs a Hermite polynomial expansion of the reduced model. This method allows us to approximate the statistics accurately with even less available data. Both methods are non-intrusive and require no a priori information of the sparsity of the system. We demonstrate the effectiveness of these two methods (Algorithms 4 and 5) using problems with up to 500 random dimensions.

Yang, Xiu↗

Solar Cycle Fine Structure and Surface Rotation from Ca II K-Line Time Series Data

Analysis of three and a half decades of data from the NSO/AFRL/Sac Peak K-line monitoring program yields evidence for four components to the variation: (a) the solar cycle, with considerable fine structure and a quasi-periodicity of 122.4 days; (b) a stochastic process, faster than (a) and largely independent of it, (c) a quasi-periodic signal due to rotational modulation, and of course (d) observational errors (shown to be quite small). Correlation and power spectrum analyses elucidate periodic and aperiodic variation of these chromospheric parameters. Time-frequency analysis is especially useful for extracting information about differential rotation, and in particular elucidates the connection between its behavior and fine structure of the solar cycle on approximately one-year time scales. These results further suggest that similar analyses will be useful at detecting and characterizing differential rotation in stars from stellar light-curves such as those being produced at NASA's Kepler observatory.

Scargle, Jeff↗

A Predictor-Corrector Strategy for Adaptivity in Dynamical Low-Rank Approximations

Here, in this paper, we present a predictor-corrector strategy for constructing rank-adaptive, dynamical low-rank approximations (DLRAs) of matrix-valued ODE systems. The strategy is a compromise between (i) low-rank step-truncation approaches that alternately evolve and compress solutions and (ii) strict DLRA approaches that augment the low-rank manifold using subspaces generated locally in time by the DLRA integrator. The strategy is based on an analysis of the error between a forward temporal update into the ambient full-rank space, which is typically computed in a step-truncation approach before recompressing, and the standard DLRA update, which is forced to live in a low-rank manifold. We use this error, without requiring its full-rank representation, to correct the DLRA solution. A key ingredient for maintaining a low-rank representation of the error is a randomized SVD, which introduces some degree of stochastic variability into the implementation. The strategy is formulated and implemented in the context of discontinuous Galerkin spatial discretizations of PDEs and applied to several versions of DLRA methods found in the literature as well as a new variant. Numerical experiments comparing the predictor-corrector strategy to other methods demonstrate robustness to overcome shortcomings of step truncation or strict DLRA approaches: The former may require more memory than is strictly needed, while the latter may miss transients solution features that cannot be recovered. The effect of randomization, tolerances, and other implementation parameters is also explored.

97 MATHEMATICS AND COMPUTING↗

Current observational constraints to the holographic dark energy model with a new infrared cutoff via the Markov chain Monte Carlo method

In this paper, the holographic dark energy models with the new infrared cutoff for both the flat case and the nonflat case are confronted with the combined constraints of current cosmological observations: type Ia supernovae, baryon acoustic oscillations, current cosmic microwave background, and the observational Hubble data. By utilizing the Markov chain Monte Carlo method, we obtain the best fit values of the parameters with 1{sigma}, 2{sigma} errors in the flat model: {Omega}{sub b}h{sup 2}=0.0230{sub -0.0010-0.0014}{sup +0.0008+0.0012}, {alpha}=0.9788{sub -0.0927-0.1249}{sup +0.1297+0.1354}, {beta}=0.4739{sub -0.0723-0.0984}{sup +0.0793+0.1055}, {Omega}{sub de0}=0.7869{sub -0.0304-0.0455}{sup +0.0291+0.0370}, {Omega}{sub m0}=0.2131{sub -0.0291-0.0370}{sup +0.0304+0.0455}, H{sub 0}=70.46{sub -2.97-4.02}{sup +2.82+3.89}. In the nonflat model, the constraint results are found in 1{sigma}, 2{sigma} regions: {Omega}{sub b}h{sup 2}=0.0229{sub -0.0010-0.0014}{sup +0.0010+0.0013}, {Omega}{sub k}=0.0014{sub -0.0597-0.0743}{sup +0.0604+0.0604}, {alpha}=0.9637{sub -0.2840-0.3333}{sup +0.2291+0.2894}, {beta}=0.4712{sub -0.0756-0.0961}{sup +0.1412+0.1703}, {Omega}{sub de0}=0.7829{sub -0.2130-0.2386}{sup +0.1588+0.1901}, {Omega}{sub m0}=0.2157{sub -0.1012-0.1241}{sup +0.1562+0.1800}, H{sub 0}=70.64{sub -3.32-4.45}{sup +2.74+3.52}. In the best fit holographic dark energy models, the equation of state of dark energy and the deceleration parameter at present are characterized by w{sub de0}=-0.9278{+-}0.0626, q{sub 0}=-0.5951{+-}0.0586 (flat case) and w{sub de0}=-0.9501{+-}0.1442, q{sub 0}=-0.6164{+-}0.0805 (nonflat case). Compared to the {Lambda}CDM model, it is found the current combined data sets do not favor the holographic dark energy model over the {Lambda}CDM model.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Reaction rates for mesoscopic reaction-diffusion kinetics

The mesoscopic reaction-diffusion master equation (RDME) is a popular modeling framework frequently applied to stochastic reaction-diffusion kinetics in systems biology. The RDME is derived from assumptions about the underlying physical properties of the system, and it may produce unphysical results for models where those assumptions fail. In that case, other more comprehensive models are better suited, such as hard-sphere Brownian dynamics (BD). Although the RDME is a model in its own right, and not inferred from any specific microscale model, it proves useful to attempt to approximate a microscale model by a specific choice of mesoscopic reaction rates. In this paper we derive mesoscopic scale-dependent reaction rates by matching certain statistics of the RDME solution to statistics of the solution of a widely used microscopic BD model: the Smoluchowski model with a Robin boundary condition at the reaction radius of two molecules. We also establish fundamental limits on the range of mesh resolutions for which this approach yields accurate results and show both theoretically and in numerical examples that as we approach the lower fundamental limit, the mesoscopic dynamics approach the microscopic dynamics. Finally, we show that for mesh sizes below the fundamental lower limit, results are less accurate. Thus, the lower limit determines the mesh size for which we obtain the most accurate results.

74 ATOMIC AND MOLECULAR PHYSICS↗

Stochastic Fermi acceleration in solar flares

Proton spectra, valid from non- to ultra-relativistic energies, resulting from stochastic Fermi acceleration in solar flares are calculated. These spectra were obtained by numerically solving the Fokker-Planck equation, in which the escape of the particles from the acceleration region is characterized by an energy-independent escape time. In addition to equilibrium spectra, time-dependent energy spectra showing the approach to equilibrium are also presented. These numerical equilibrium spectra are compared with previous results which were obtained either by Monte Carlo simulations or approximate analytical treatments. There are no analytic solutions valid in the transrelativistic regime, which is very important for the production of pions and neutrons in solar flares. The acceleration efficiency is related to physical parameters, in particular the energy density in either magnetosonic or Alfven waves, and a lower limit is placed on either of these energy densities from acceleration times implied by gamma-ray observations. Also discussed is the physical interpretation of the escape time.

Miller, James A.↗

Physical Models of Layered Polar Firn Brightness Temperatures from 0.5 to 2 GHz

We investigate physical effects influencing 0.5-2 GHz brightness temperatures of layered polar firn to support the Ultra Wide Band Software Defined Radiometer (UWBRAD) experiment to be conducted in Greenland and in Antarctica. We find that because ice particle grain sizes are very small compared to the 0.5-2 GHz wavelengths, volume scattering effects are small. Variations in firn density over cm- to m-length scales, however, cause significant effects. Both incoherent and coherent models are used to examine these effects. Incoherent models include a 'cloud model' that neglects any reflections internal to the ice sheet, and the DMRT-ML and MEMLS radiative transfer codes that are publicly available. The coherent model is based on the layered medium implementation of the fluctuation dissipation theorem for thermal microwave radiation from a medium having a nonuniform temperature. Density profiles are modeled using a stochastic approach, and model predictions are averaged over a large number of realizations to take into account an averaging over the radiometer footprint. Density profiles are described by combining a smooth average density profile with a spatially correlated random process to model density fluctuations. It is shown that coherent model results after ensemble averaging depend on the correlation lengths of the vertical density fluctuations. If the correlation length is moderate or long compared with the wavelength (approximately 0.6x longer or greater for Gaussian correlation function without regard for layer thinning due to compaction), coherent and incoherent model results are similar (within approximately 1 K). However, when the correlation length is short compared to the wavelength, coherent model results are significantly different from the incoherent model by several tens of kelvins. For a 10-cm correlation length, the differences are significant between 0.5 and 1.1 GHz, and less for 1.1-2 GHz. Model results are shown to be able to match the v-pol SMOS data closely and predict the h-pol data for small observation angles.

layered polar firn emission↗

Analysis of Monte Carlo accelerated iterative methods for sparse linear systems: Analysis of Monte Carlo accelerated iterative methods for sparse linear systems

Here, we consider hybrid deterministic-stochastic iterative algorithms for the solution of large, sparse linear systems. Starting from a convergent splitting of the coefficient matrix, we analyze various types of Monte Carlo acceleration schemes applied to the original preconditioned Richardson (stationary) iteration. We expect that these methods will have considerable potential for resiliency to faults when implemented on massively parallel machines. We also establish sufficient conditions for the convergence of the hybrid schemes, and we investigate different types of preconditioners including sparse approximate inverses. Numerical experiments on linear systems arising from the discretization of partial differential equations are presented.

97 MATHEMATICS AND COMPUTING↗

Stochastic Filtering of Rain Profiles Using Radar, Surface-Referenced Radar, or Combined Radar-Radiometer Measurements

This paper describes a computationally efficient nearly optimal Bayesian algorithm to estimate rain (and drop size distribution) profiles, given a radar reflectivity profile at a single attenuating wavelength. In addition to estimating the averages of all the mutually ambiguous combinations of rain parameters that can produce the data observed, the approach also calculates the n-ns uncertainty in its estimates (this uncertainty thus quantifies "the amount of ambiguity" in the "solution"). The paper also describes a more general approach that can make estimates based on a radar reflectivity profile together with an approximate measurement of the path-integrated attenuation, or a radar reflectivity profile and a set of passive microwave brightness temperatures. This more general "combined" algorithm is currently being adapted for the Tropical Rainfall Measuring Mission.

Haddad, Ziad S.↗

POD Model Reconstruction for Gray-Box Fault Detection

Proper orthogonal decomposition (POD) is the mathematical basis of a method of constructing low-order mathematical models for the "gray-box" fault-detection algorithm that is a component of a diagnostic system known as beacon-based exception analysis for multi-missions (BEAM). POD has been successfully applied in reducing computational complexity by generating simple models that can be used for control and simulation for complex systems such as fluid flows. In the present application to BEAM, POD brings the same benefits to automated diagnosis. BEAM is a method of real-time or offline, automated diagnosis of a complex dynamic system.The gray-box approach makes it possible to utilize incomplete or approximate knowledge of the dynamics of the system that one seeks to diagnose. In the gray-box approach, a deterministic model of the system is used to filter a time series of system sensor data to remove the deterministic components of the time series from further examination. What is left after the filtering operation is a time series of residual quantities that represent the unknown (or at least unmodeled) aspects of the behavior of the system. Stochastic modeling techniques are then applied to the residual time series. The procedure for detecting abnormal behavior of the system then becomes one of looking for statistical differences between the residual time series and the predictions of the stochastic model.

Park, Han↗