Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “stochastic models”

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 1,783 records · Page 99

Application of remote sensing to hydrology

Streamflow forecasting and hydrologic modelling are considered in a feasibility assessment of using the data produced by remote observation from space and/or aircraft to reduce the time and expense normally involved in achieving the ability to predict the hydrological behavior of an ungaged watershed. Existing watershed models are described, and both stochastic and parametric techniques are discussed towards the selection of a suitable simulation model. Technical progress and applications are reported and recommendations are made for additional research.

Ambaruch, R.↗

On Parallelizing Single Dynamic Simulation Using HPC Techniques and APIs of Commercial Software

Time-domain simulations are heavily used in today’s planning and operation practices to assess power system transient stability and post-transient voltage/frequency profiles following severe contingencies to comply with industry standards. Because of the increased modeling complexity, it is several times slower than real time for state-of-the-art commercial packages to complete a dynamic simulation for a large-scale model. With the growing stochastic behavior introduced by emerging technologies, power industry has seen a growing need for performing security assessment in real time. This paper presents a parallel implementation framework to speed up a single dynamic simulation by leveraging the existing stability model library in commercial tools through their application programming interfaces (APIs). Several high performance computing (HPC) techniques are explored such as parallelizing the calculation of generator current injection, identifying fast linear solvers for network solution, and parallelizing data outputs when interacting with APIs in the commercial package, TSAT. The proposed method has been tested on a WECC planning base case with detailed synchronous generator models and exhibits outstanding scalable performance with sufficient accuracy.

Diao, Ruisheng↗

Thermal fluctuations in the Landau-Lifshitz-Bloch model

A formulation for thermal noise in the stochastic form of the Landau-Lifshitz-Bloch equation used for modeling the magnetization dynamics at elevated temperatures is presented here. The diffusion coefficients for thermal fluctuations are obtained via the Fokker-Plank equation using the mean-field approximation of the field in defining the free energy. The presented model leads to a mean magnetization consistent with the equilibrium magnetization for small and large particles. The distribution of the magnetization magnitude is of the Maxwell-Boltzmann type. The presented model was tested by studying the equilibrium magnetization in macrospin particles at high temperatures. The model is appealing for multiscale modeling, such as modeling heat assisted magnetic recording systems and all-optical magnetization reversal.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Influence of spatial variability of hydraulic characteristics of soils on surface parameters obtained from remote sensing data in infrared and microwaves

The correct interpretation of thermal and hydraulic soil parameters infrared from remotely sensed data (thermal infrared, microwaves) implies a good understanding of the causes of their temporal and spatial variability. Given this necessity, the sensitivity of the surface variables (temperature, moisture) to the spatial variability of hydraulic soil properties is tested with a numerical model of heat and mass transfer between bare soil and atmosphere. The spatial variability of hydraulic soil properties is taken into account in terms of the scaling factor. For a given soil, the knowledge of its frequency distribution allows a stochastic use of the model. The results are treated statistically, and the part of the variability of soil surface parameters due to that of soil hydraulic properties is evaluated quantitatively.

Brunet, Y.↗

Development of neural network convection parameterizations for climate and NWP models using Cloud Resolving Model simulations

A novel approach based on the neural network (NN) technique is formulated and used for development of a NN ensemble stochastic convection parameterization for climate models. This fast parameterization is built based on data from Cloud Resolving Model (CRM) simulations initialized with and driven by the TOGA-COARE data available for the 4-month winter season from November 1992 to February 1993. CRM simulated data were averaged and projected onto the GCM space of atmospheric states to implicitly define a stochastic convection parameterization. This parameterization is emulated using an ensemble of neural networks (NNs). The developed NNs are trained and tested. The inherent uncertainty of the stochastic convection parameterization derived following this approach is estimated. The newly developed NN convection parameterization has been tested in a diagnostic mode of NCAR CAM. It produced reasonable and promising climate simulations for the TOGA-COARE winter. This paper is devoted to discussion of the concept, methodology, initial results,and the major challenges of development of NN convection parameterizations for climate and numerical weather prediction (NWR) models.

97 MATHEMATICS AND COMPUTING↗

An Adaptive Sparse-Grid High-Order Stochastic Collocation Method for Bayesian Inference in Groundwater Reactive Transport Modeling

Although Bayesian analysis has become vital to the quantification of prediction uncertainty in groundwater modeling, its application has been hindered due to the computational cost associated with numerous model executions needed for exploring the posterior probability density function (PPDF) of model parameters. This is particularly the case when the PPDF is estimated using Markov Chain Monte Carlo (MCMC) sampling. In this study, we develop a new approach that improves computational efficiency of Bayesian inference by constructing a surrogate system based on an adaptive sparse-grid high-order stochastic collocation (aSG-hSC) method. Unlike previous works using first-order hierarchical basis, we utilize a compactly supported higher-order hierar- chical basis to construct the surrogate system, resulting in a significant reduction in the number of computational simulations required. In addition, we use hierarchical surplus as an error indi- cator to determine adaptive sparse grids. This allows local refinement in the uncertain domain and/or anisotropic detection with respect to the random model parameters, which further improves computational efficiency. Finally, we incorporate a global optimization technique and propose an iterative algorithm for building the surrogate system for the PPDF with multiple significant modes. Once the surrogate system is determined, the PPDF can be evaluated by sampling the surrogate system directly with very little computational cost. The developed method is evaluated first using a simple analytical density function with multiple modes and then using two synthetic groundwater reactive transport models. The groundwater models represent different levels of complexity; the first example involves coupled linear reactions and the second example simulates nonlinear ura- nium surface complexation. The results show that the aSG-hSC is an effective and efficient tool for Bayesian inference in groundwater modeling in comparison with conventional MCMC sim- ulations. The computational efficiency is expected to be more beneficial to more computational expensive groundwater problems.

Zhang, Guannan↗

A data-driven framework for sparsity-enhanced surrogates with arbitrary mutually dependent randomness

The challenge of quantifying uncertainty propagation in real-world systems is rooted in the high-dimensionality of the stochastic input and the frequent lack of explicit knowledge of its probability distribution. Traditional approaches show limitations for such problems, especially when the size of the training data is limited. To address these diculties, we have developed a general framework of constructing surrogate models on spaces of stochastic input with arbitrary probability measure irrespective of the mutual dependencies between individual components of the random inputs and the analytical form. The present Data-driven Sparsity-enhancing Rotation for Arbitrary Randomness (DSRAR) framework includes a data-driven construction of multivariate polynomial basis for arbitrary mutually dependent probability measure and a sparsity enhancement rotation procedure. This sparsity-enhancing rotation method was initially proposed in our previous work [1] for Gaussian density distributions, which may not be feasible for non-Gaussian distributions due to the loss of orthogonality after the rotation. To remedy such diculties, we developed a new data-driven approach to construct orthonormal polynomials for polynomials for arbitrary mutually dependent (amdP) randomness, ensuring the constructed basis maintains the orthogonality/near-orthogonality with respect to the density of the rotated random vector, where directly applying the regular polynomial chaos including arbitrary polynomial chaos (aPC) [2] shows limitations due to the assumption of the mutual independence between the components of the random inputs. The developed DSRAR framework leads to accurate recovery, with only limited training data, of a sparse representation of the target functions. The eectiveness of our method is demonstrated in challenging problems such as PDEs and realistic molecular systems within high-dimensional conformational space (O(10)) where the underlying density is implicitly represented by a large collection of sample data, as well as systems with explicitly given non-Gaussian probabilistic measures.

Lei, Huan↗

Spatio‐temporally Resolved Methane Fluxes From the Los Angeles Megacity

Abstract We combine sustained observations from a network of atmospheric monitoring stations with inverse modeling to uniquely obtain spatiotemporal (3‐km, 4‐day) estimates of methane emissions from the Los Angeles megacity and the broader South Coast Air Basin for 2015–2016. Our inversions use customized and validated high‐fidelity meteorological output from Weather Research Forecasting and Stochastic Time‐Inverted Lagrangian model for South Coast Air Basin and innovatively employ a model resolution matrix‐based metric to disentangle the spatiotemporal information content of observations as manifested through estimated fluxes. We partially track and constrain fluxes from the Aliso Canyon natural gas leak and detect closure of the Puente Hills landfill, with no prior information. Our annually aggregated fluxes and their uncertainty excluding the Aliso Canyon leak period lie within the uncertainty bounds of the fluxes reported by the previous studies. Spatially, major sources of CH 4 emissions in the basin were correlated with CH 4 ‐emitting infrastructure. Temporally, our findings show large seasonal variations in CH 4 fluxes with significantly higher fluxes in winter in comparison to summer months, which is consistent with natural gas demand and anticorrelated with air temperature. Overall, this is the first study that utilizes inversions to detect both enhancement (Aliso Canyon leak) and reduction (Puente Hills) in CH 4 fluxes due to the unintended events and policy decisions and thereby demonstrates the utility of inverse modeling for identifying variations in fluxes at fine spatiotemporal resolution.

Yadav, Vineet↗

Fluctuation and dissipation dynamics in fusion reactions from a stochastic mean-field approach

By projecting the stochastic mean-field dynamics on a suitable collective path during the entrance channel of heavy-ion collisions, expressions for transport coefficients associated with relative distance are extracted. These transport coefficients, which have forms similar to those familiar from the nucleon exchange model, are evaluated by carrying out time-dependent Hartree-Fock simulations. The calculations provide an accurate description of the magnitude and form factor of transport coefficients associated with one-body dissipation and fluctuation mechanisms.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Realization of consensus of multi-agent systems with stochastically mixed interactions

In this paper, we propose a new consensus model in which the interactions among agents stochastically switch between attraction and repulsion. Such a positive-and-negative mechanism is described by the white-noise-based coupling. Analytic criteria for the consensus and non-consensus in terms of the eigenvalues of the noise intensity matrix are derived, which provide a better understanding of the constructive roles of random interactions. Specifically, we discover a positive role of noise coupling that noise can accelerate the emergence of consensus. We find that the converging speed of the multi-agent network depends on the square of the second smallest eigenvalue of its graph Laplacian. The influence of network topologies on the consensus time is also investigated.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Optimal Proactive Vehicle Relocation for On-Demand Mobility Service with Deep Convolution-LSTM Network

This study develops an innovative model for inferring optimal proactive relocation strategy for empty vehicles of on-demand mobility service (OMS) directly from past observations of passenger demand. The output of our model is a row stochastic relocation matrix which gives the probability that an empty vehicle should be relocated from location i to location j at any given time. The matrix itself is high-dimensional and therefore challenging to obtain accurate estimation with traditional machine learning approach. Our model proposes the low-rank approximation of the relocation matrix, and introduces the deep convolution-LSTM network which extracts low-level spatial and temporal features from past demand observations for best restoration of the actual low rank approximation vector. The model is trained by deriving the optimal matching strategies from one year of real-world trip data. The results show that our model is able to predict the low-rank approximation of the relocation matrix with 4% of mean absolute error, may reduce the relocation cost by 20%-45% as compared to the real-world observations and is 50% optimal as compared to the global optimal matching with perfect information.

Lei, Zengxiang↗

Variability and Spectral Characteristics of Three Flaring Gamma-Ray Quasars Observed by VERITAS and Fermi-LAT

Flat-spectrum radio quasars (FSRQs) are the most luminous blazars at GeV energies but only rarely emit detectable fluxes of TeV gamma rays, typically during bright GeV flares. We explore the gamma-ray variability and spectral characteristics of three FSRQs that have been observed at GeV and TeV energies by Fermi-LAT and VERITAS, making use of almost 100 hr of VERITAS observations spread over 10 yr: 3C 279, PKS 1222+216, and Ton 599. We explain the GeV flux distributions of the sources in terms of a model derived from a stochastic differential equation describing fluctuations in the magnetic field in the accretion disk and estimate the timescales of magnetic flux accumulation and stochastic instabilities in their accretion disks. We identify distinct flares using a procedure based on Bayesian blocks and analyze their daily and subdaily variability and gamma-ray energy spectra. Using observations from VERITAS, as well as Fermi, Swift, and the Steward Observatory, we model the broadband spectral energy distributions of PKS 1222+216 and Ton 599 during very high energy (VHE)–detected flares in 2014 and 2017, respectively, strongly constraining the jet Doppler factors and gamma-ray emission region locations during these events. Finally, we place theoretical constraints on the potential production of PeV-scale neutrinos during these VHE flares.

79 ASTRONOMY AND ASTROPHYSICS↗

Multiscale eddy simulation for moist atmospheric convection: Preliminary investigation

A multiscale computational framework is designed for simulating atmospheric convection and clouds. In this multiscale framework, large eddy simulation (LES) is used to model the coarse scales of 100 m and larger, and a stochastic, one-dimensional turbulence (ODT) model is used to represent the fine scales of 100 m and smaller. Coupled and evolving together, these two components provide a multiscale eddy simulation (MES). Through its fine-scale turbulence and moist thermodynamics, MES allows coarse grid cells to be partially cloudy and to encompass cloudy–clear air mixing on scales down to 1 m; in contrast, in typical LES such fine-scale processes are not represented or are parameterized using bulk deterministic closures. To illustrate MES and investigate its multiscale dynamics, a shallow cumulus cloud field is simulated. The fine-scale variability is seen to take a plausible form, with partially cloudy grid cells prominent near cloud edges and cloud top. From earlier theoretical work, this mixing of cloudy and clear air is believed to have an important impact on buoyancy. However, contrary to expectations based on earlier theoretical studies, the mean statistics of the bulk cloud field are essentially the same in MES and LES; possible reasons for this are discussed, including possible limitations in the present formulation of MES. One difference between LES and MES is seen in the coarse-scale turbulent kinetic energy, which appears to grow slowly in time due to incoherent stochastic fluctuations in the buoyancy. This and other considerations suggest the need for some type of spatial and/or temporal filtering to attenuate undersampling of the stochastic fine-scale processes.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Diffusion of two-dimensional epitaxial clusters on metal (100) surfaces: Facile versus nucleation-mediated behavior and their merging for larger sizes

For diffusion of two-dimensional homoepitaxial clusters of N atoms on metal(100) surfaces mediated by edge atom hopping, macroscale continuum theory suggests that the diffusion coefficient scales like DN ~ N -β with β = 3/2. However, we find quite different and diverse behavior in multiple size regimes. These include: (i) facile diffusion for small sizes N < 9; (ii) slow nucleation-mediated diffusion with small β < 1 for “perfect” sizes N = N p = L 2 or L(L+1), for L = 3, 4,… having unique ground state shapes, for moderate sizes 9 ≤ N ≤ O(10 2 ); the same also applies for N = N p +3, N p + 4,… (iii) facile diffusion but with large β > 2 for N = Np + 1 and N p + 2 also for moderate sizes 9 ≤ N ≤ O(10 2 ); (iv) merging of the above distinct branches and subsequent anomalous scaling with 1 ≲ β < 3/2, reflecting the quasi-facetted structure of clusters, for larger N = O(10 2 ) to N = O(10 3 ); and (v) classic scaling with β = 3/2 for very large N = O(103) and above. The specified size ranges apply for typical model parameters. We focus on the moderate size regime where show that diffusivity cycles quasi-periodically from the slowest branch for N p + 3 (not Np) to the fastest branch for Np + 1. Behavior is quantified by Kinetic Monte Carlo simulation of an appropriate stochastic lattice-gas model. However, precise analysis must account for a strong enhancement of diffusivity for short time increments due to back-correlation in the cluster motion. Further understanding of this enhancement, of anomalous size scaling behavior, and of the merging of various branches, is facilitated by combinatorial analysis of the number of the ground state and low-lying excited state cluster configurations, and also of kink populations.

36 MATERIALS SCIENCE↗

Accurate modeling and reconstruction of three-dimensional percolating filamentary microstructures from two-dimensional micrographs via dilation-erosion method

Heterogeneous materials are ubiquitous in nature and synthetic situations and have a wide range of important engineering applications. Accurate modeling and reconstructing three-dimensional (3D) microstructure of topologically complex materials from limited morphological information such as a two-dimensional (2D) micrograph is crucial to the assessment and prediction of effective material properties and performance under extreme conditions. Here, we extend a recently developed dilation–erosion method and employ the Yeong–Torquato stochastic reconstruction procedure to model and generate 3D austenitic–ferritic cast duplex stainless steel microstructure containing percolating filamentary ferrite phase from 2D optical micrographs of the material sample. Specifically, the ferrite phase is dilated to produce a modified target 2D microstructure and the resulting 3D reconstruction is eroded to recover the percolating ferrite filaments. The dilation–erosion reconstruction is compared with the actual 3D microstructure, obtained from serial sectioning (polishing), as well as the standard stochastic reconstructions incorporating topological connectedness information. The fact that the former can achieve the same level of accuracy as the latter suggests that the dilation–erosion procedure is tantamount to incorporating appreciably more topological and geometrical information into the reconstruction while being much more computationally efficient. - Highlights: • Spatial correlation functions used to characterize filamentary ferrite phase • Clustering information assessed from 3D experimental structure via serial sectioning • Stochastic reconstruction used to generate 3D virtual structure 2D micrograph • Dilation–erosion method to improve accuracy of 3D reconstruction.

36 MATERIALS SCIENCE↗

Computational modeling of multispectral remote sensing systems: Background investigations

A computational model of the deterministic and stochastic process of remote sensing has been developed based upon the results of the investigations presented. The model is used in studying concepts for improving worldwide environment and resource monitoring. A review of various atmospheric radiative transfer models is presented as well as details of the selected model. Functional forms for spectral diffuse reflectance with variability introduced are also presented. A cloud detection algorithm and the stochastic nature of remote sensing data with its implications are considered.

Aherron, R. M.↗

Radio Occultation Investigation of the Rings of Saturn and Uranus

The proposed work addresses two main objectives: (1) to pursue the development of the random diffraction screen model for analytical/computational characterization of the extinction and near-forward scattering by ring models that include particle crowding, uniform clustering, and clustering along preferred orientations (anisotropy). The characterization is crucial for proper interpretation of past (Voyager) and future (Cassini) ring, occultation observations in terms of physical ring properties, and is needed to address outstanding puzzles in the interpretation of the Voyager radio occultation data sets; (2) to continue the development of spectral analysis techniques to identify and characterize the power scattered by all features of Saturn's rings that can be resolved in the Voyager radio occultation observations, and to use the results to constrain the maximum particle size and its abundance. Characterization of the variability of surface mass density among the main ring, features and within individual features is important for constraining the ring mass and is relevant to investigations of ring dynamics and origin. We completed the developed of the stochastic geometry (random screen) model for the interaction of electromagnetic waves with of planetary ring models; used the model to relate the oblique optical depth and the angular spectrum of the near forward scattered signal to statistical averages of the stochastic geometry of the randomly blocked area. WE developed analytical results based on the assumption of Poisson statistics for particle positions, and investigated the dependence of the oblique optical depth and angular spectrum on the fractional area blocked, vertical ring profile, and incidence angle when the volume fraction is small. Demonstrated agreement with the classical radiative transfer predictions for oblique incidence. Also developed simulation procedures to generate statistical realizations of random screens corresponding to uniformly packed ring models, and used the results to characterize dependence of the extinction and near-forward scattering on ring thickness, packing fraction, and the ring opening angle.

Marouf, Essam A.↗

Learning Stochastic Parametric Differentiable Predictive Control Policies

We present a scalable unsupervised learning-based method for obtaining explicit control policies for model predictive control problems for stochastic linear systems with additive uncertainties subject to nonlinear chance constraints. We call the proposed method stochastic parametric differentiable predictive control (SP-DPC), which extends the recently proposed deterministic DPC policy optimization algorithm. We formulate the SP-DPC as a deterministic approximation to the stochastic parametric constrained optimal control problem via independent sampling of the problem's parameters and uncertainties. This formulation allows us to directly compute the policy gradients via automatic differentiation of the problem's value function, evaluated over sampled parameters and uncertainties. In particular, the computed expectation of the problem's value function is backpropagated through the finite-time closed-loop system rollouts parametrized by a known nominal system dynamics model and neural control policy. We also provide theoretical probabilistic guarantees on closed-loop stability and chance constraints satisfaction for systems controlled by learned neural policies. We demonstrate the computational efficiency and scalability of the proposed policy optimization algorithm in three numerical examples, including systems with a large number of states or subject to nonlinear constraints.

Drgona, Jan↗