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 1,369 records · Page 76
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Retrieving microphysical properties of concurrent pristine ice and snow using polarimetric radar observations

Abstract. Ice and mixed-phase clouds play a key role in our climate system because of their strong controls on global precipitation and radiation budget. Their microphysical properties have been characterized commonly by polarimetric radar measurements. However, there remains a lack of robust estimates of microphysical properties of concurrent pristine ice and aggregates because larger snow aggregates often dominate the radar signal and mask contributions of smaller pristine ice crystals. This paper presents a new method that separates the scattering signals of pristine ice embedded in snow aggregates in scanning polarimetric radar observations and retrieves their respective abundances and sizes for the first time. This method, dubbed ENCORE-ice, is built on an iterative stochastic ensemble retrieval framework. It provides the number concentration, ice water content, and effective mean diameter of pristine ice and snow aggregates with uncertainty estimates. Evaluations against synthetic observations show that the overall retrieval biases in the combined total microphysical properties are within 5 % and that the errors with respect to the truth are well within the retrieval uncertainty. The partitioning between pristine ice and snow aggregates also agrees well with the truth. Additional evaluations against in situ cloud probe measurements from a recent campaign for a stratiform cloud system are promising. Our median retrievals have a bias of 98 % in the total ice number concentration and 44 % in the total ice water content. This performance is generally better than the retrieval from empirical relationships. The ability to separate signals of different ice species and to provide their quantitative microphysical properties will open up many research opportunities, such as secondary ice production studies and model evaluations for ice microphysical processes.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1985

Implicit dual control for general stochastic systems

A new implicit dual technique is presented for stochastic adaptive control synthesis. In this technique, denoted as the method of utility costs (MOUC), a new control policy is derived from a known starting control policy by using a single iteration in time-varying policy space. An important result is that the new control policy yields a corresponding improvement in adaptive performance relative to the known starting policy. Since the starting policy can be chosen arbitrarily, the new policy has potentially superior performance relative to any existing design. Unlike other implicit dual designs, the MOUC can be applied to systems having constrained inputs, discontinuities, discrete valued disturbances, as well as in situations where the variances of the associated process statistics are large. A numerical example is given in which a controller is synthesized to provide improved performance relative to the Open-Loop Feedback policy of Dreyfus.

Bayard, D. S.↗
NASA NTRS1993

Estimation of Kalman filter gain from output residuals

This paper presents a procedure to estimate the Kalman filter gain from input-output measurement data with a given system model. The system model can be a finite element model or an experimental model from any identification method. The procedure consists of three basic steps. First, the stochastic portion related to the residuals of the response is computed. Second, the coefficients of a linear difference model for the stochastic portion are estimated by a least-squares solution that minimizes the filter residual. Third, the Kalman filter gain is computed from these model coefficients. Experimental results are presented to illustrate the usefulness of the developed procedure.

Juang, Jer-Nan↗
NASA NTRS2009

3D Lunar Terrain Reconstruction from Apollo Images

Generating accurate three dimensional planetary models is becoming increasingly important as NASA plans manned missions to return to the Moon in the next decade. This paper describes a 3D surface reconstruction system called the Ames Stereo Pipeline that is designed to produce such models automatically by processing orbital stereo imagery. We discuss two important core aspects of this system: (1) refinement of satellite station positions and pose estimates through least squares bundle adjustment; and (2) a stochastic plane fitting algorithm that generalizes the Lucas-Kanade method for optimal matching between stereo pair images.. These techniques allow us to automatically produce seamless, highly accurate digital elevation models from multiple stereo image pairs while significantly reducing the influence of image noise. Our technique is demonstrated on a set of 71 high resolution scanned images from the Apollo 15 mission

Broxton, Michael J.↗
NASA NTRS2019

Sizing and Topology Design of an Aeroelastic Wingbox Under Uncertainty

The goals of this work are to use a nested optimizer to conduct simultaneous sizing (inner level) and topology (outer level) design of a wingbox, considering uncertainties in the safety factors used to define the aeroelastic constraints. These uncertainties, propagated via sampling-driven polynomial chaos, are explicitly introduced at the inner level of the method, during gradient-based sizing optimization, resulting in a stochastic optimal sizing distribution. Measures of robustness in the total structural mass are then passed to the outer level, where a global optimizer evolves the topology parameters. The results demonstrate design choices needed to improve robustness in the face of uncertain safety factors, and the various physical mechanisms driving this process.

Stanford, Bret K.↗
DOE OSTI2020

Nonlinear power flow control for networked AC/DC microgrids

A method for designing feedforward and feedback controllers for integration of stochastic sources and loads into a nonlinear networked AC/DC microgrid system is provided. A reduced order model for general networked AC/DC microgrid systems is suitable for HSSPFC control design. A simple feedforward steady state solution is utilized for the feedforward controls block. Feedback control laws are provided for the energy storage systems. A HSSPFC controller design is implemented that incorporates energy storage systems that provides static and dynamic stability conditions for both the DC random stochastic input side and the AC random stochastic load side. Transient performance was investigated for the feedforward/feedback control case. Numerical simulations were performed and provided power and energy storage profile requirements for the networked AC/DC microgrid system overall performance. The HSSPFC design can be implemented in the Matlab/Simulink environment that is compatible with real time simulation/controllers.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2017

Dynamic Price Vector Formation Model-Based Automatic Demand Response Strategy for PV-Assisted EV Charging Stations

A real-time price (RTP)-based automatic demand response (ADR) strategy for PV-assisted electric vehicle (EV) Charging Station (PVCS) without vehicle to grid is proposed. The charging process is modeled as a dynamic linear program instead of the normal day-ahead and real-time regulation strategy, to capture the advantages of both global and real-time optimization. Different from conventional price forecasting algorithms, a dynamic price vector formation model is proposed based on a clustering algorithm to form an RTP vector for a particular day. A dynamic feasible energy demand region (DFEDR) model considering grid voltage profiles is designed to calculate the lower and upper bounds. A deduction method is proposed to deal with the unknown information of future intervals, such as the actual stochastic arrival and departure times of EVs, which make the DFEDR model suitable for global optimization. Finally, both the comparative cases articulate the advantages of the developed methods and the validity in reducing electricity costs, mitigating peak charging demand, and improving PV self-consumption of the proposed strategy are verified through simulation scenarios.

14 SOLAR ENERGY↗
DOE OSTI2016

Concurrent Monte Carlo transport and fluence optimization with fluence adjusting scalable transport Monte Carlo

Purpose: The future of radiation therapy will require advanced inverse planning solutions to support single-arc, multiple-arc, and “4π” delivery modes, which present unique challenges in finding an optimal treatment plan over a vast search space, while still preserving dosimetric accuracy. The successful clinical implementation of such methods would benefit from Monte Carlo (MC) based dose calculation methods, which can offer improvements in dosimetric accuracy when compared to deterministic methods. The standard method for MC based treatment planning optimization leverages the accuracy of the MC dose calculation and efficiency of well-developed optimization methods, by precalculating the fluence to dose relationship within a patient with MC methods and subsequently optimizing the fluence weights. However, the sequential nature of this implementation is computationally time consuming and memory intensive. Methods to reduce the overhead of the MC precalculation have been explored in the past, demonstrating promising reductions of computational time overhead, but with limited impact on the memory overhead due to the sequential nature of the dose calculation and fluence optimization. The authors propose an entirely new form of “concurrent” Monte Carlo treat plan optimization: a platform which optimizes the fluence during the dose calculation, reduces wasted computation time being spent on beamlets that weakly contribute to the final dose distribution, and requires only a low memory footprint to function. In this initial investigation, the authors explore the key theoretical and practical considerations of optimizing fluence in such a manner. Methods: The authors present a novel derivation and implementation of a gradient descent algorithm that allows for optimization during MC particle transport, based on highly stochastic information generated through particle transport of very few histories. A gradient rescaling and renormalization algorithm, and the concept of momentum from stochastic gradient descent were used to address obstacles unique to performing gradient descent fluence optimization during MC particle transport. The authors have applied their method to two simple geometrical phantoms, and one clinical patient geometry to examine the capability of this platform to generate conformal plans as well as assess its computational scaling and efficiency, respectively. Results: The authors obtain a reduction of at least 50% in total histories transported in their investigation compared to a theoretical unweighted beamlet calculation and subsequent fluence optimization method, and observe a roughly fixed optimization time overhead consisting of ∼10% of the total computation time in all cases. Finally, the authors demonstrate a negligible increase in memory overhead of ∼7–8 MB to allow for optimization of a clinical patient geometry surrounded by 36 beams using their platform. Conclusions: This study demonstrates a fluence optimization approach, which could significantly improve the development of next generation radiation therapy solutions while incurring minimal additional computational overhead.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2012

AGE AND MASS FOR 920 LARGE MAGELLANIC CLOUD CLUSTERS DERIVED FROM 100 MILLION MONTE CARLO SIMULATIONS

We present new age and mass estimates for 920 stellar clusters in the Large Magellanic Cloud (LMC) based on previously published broadband photometry and the stellar cluster analysis package, MASSCLEANage. Expressed in the generic fitting formula, d{sup 2} N/dMdt{proportional_to}M{sup {alpha}} t{sup {beta}}, the distribution of observed clusters is described by {alpha} = -1.5 to -1.6 and {beta} = -2.1 to -2.2. For 288 of these clusters, ages have recently been determined based on stellar photometric color-magnitude diagrams, allowing us to gauge the confidence of our ages. The results look very promising, opening up the possibility that this sample of 920 clusters, with reliable and consistent age, mass, and photometric measures, might be used to constrain important characteristics about the stellar cluster population in the LMC. We also investigate a traditional age determination method that uses a {chi}{sup 2} minimization routine to fit observed cluster colors to standard infinite-mass limit simple stellar population models. This reveals serious defects in the derived cluster age distribution using this method. The traditional {chi}{sup 2} minimization method, due to the variation of U, B, V, R colors, will always produce an overdensity of younger and older clusters, with an underdensity of clusters in the log (age/yr) = [7.0, 7.5] range. Finally, we present a unique simulation aimed at illustrating and constraining the fading limit in observed cluster distributions that includes the complex effects of stochastic variations in the observed properties of stellar clusters.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2014

The rigorous stochastic matrix multiplication scheme for the calculations of reduced equilibrium density matrices of open multilevel quantum systems

Understanding the roles of the temporary and spatial structures of quantum functional noise in open multilevel quantum molecular systems attracts a lot of theoretical interests. I want to establish a rigorous and general framework for functional quantum noises from the constructive and computational perspectives, i.e., how to generate the random trajectories to reproduce the kernel and path ordering of the influence functional with effective Monte Carlo methods for arbitrary spectral densities. This construction approach aims to unify the existing stochastic models to rigorously describe the temporary and spatial structure of Gaussian quantum noises. In this paper, I review the Euclidean imaginary time influence functional and propose the stochastic matrix multiplication scheme to calculate reduced equilibrium density matrices (REDM). In addition, I review and discuss the Feynman-Vernon influence functional according to the Gaussian quadratic integral, particularly its imaginary part which is critical to the rigorous description of the quantum detailed balance. As a result, I establish the conditions under which the influence functional can be interpreted as the average of exponential functional operator over real-valued Gaussian processes for open multilevel quantum systems. I also show the difference between the local and nonlocal phonons within this framework. With the stochastic matrix multiplication scheme, I compare the normalized REDM with the Boltzmann equilibrium distribution for open multilevel quantum systems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

A new method to measure galaxy bias by combining the density and weak lensing fields

We present a new method to measure redshift-dependent galaxy bias by combining information from the galaxy density field and the weak lensing field. This method is based on the work of Amara et al., who use the galaxy density field to construct a bias-weighted convergence field κg. The main difference between Amara et al.'s work and our new implementation is that here we present another way to measure galaxy bias, using tomography instead of bias parametrizations. The correlation between κg and the true lensing field κ allows us to measure galaxy bias using different zero-lag correlations, such as <κgκ>/<κκ> or <κgκg>/<κgκ>. Our method measures the linear bias factor on linear scales, under the assumption of no stochasticity between galaxies and matter. We use the Marenostrum Institut de Ciències de l'Espai (MICE) simulation to measure the linear galaxy bias for a flux-limited sample (i < 22.5) in tomographic redshift bins using this method. This article is the first that studies the accuracy and systematic uncertainties associated with the implementation of the method and the regime in which it is consistent with the linear galaxy bias defined by projected two-point correlation functions (2PCF). We find that our method is consistent with a linear bias at the per cent level for scales larger than 30 arcmin, while non-linearities appear at smaller scales. This measurement is a good complement to other measurements of bias, since it does not depend strongly on σ8 as do the 2PCF measurements. We will apply this method to the Dark Energy Survey Science Verification data in a follow-up article.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Bayesian sparse learning with preconditioned stochastic gradient MCMC and its applications

Deep neural networks have been successfully employed in an extensive variety of research areas, including solving partial differential equations. Despite its significant success, there are some challenges in effectively training DNN, such as avoiding overfitting in over-parameterized DNNs and accelerating the optimization in DNNs with pathological curvature. Here, we propose a Bayesian type sparse deep learning algorithm. The algorithm utilizes a set of spike-and-slab priors for the parameters in the deep neural network. The hierarchical Bayesian mixture will be trained using an adaptive empirical method. That is, one will alternatively sample from the posterior using preconditioned stochastic gradient Langevin Dynamics (PSGLD), and optimize the latent variables via stochastic approximation. The sparsity of the network is achieved while optimizing the hyperparameters with adaptive searching and penalizing. A popular SG-MCMC approach is Stochastic gradient Langevin dynamics (SGLD). However, considering the complex geometry in the model parameter space in nonconvex learning, updating parameters using a universal step size in each component as in SGLD may cause slow mixing. To address this issue, we apply a computationally manageable preconditioner in the updating rule, which provides a step-size parameter to adapt to local geometric properties. Moreover, by smoothly optimizing the hyperparameter in the preconditioning matrix, our proposed algorithm ensures a decreasing bias, which is introduced by ignoring the correction term in the preconditioned SGLD. According to the existing theoretical framework, we show that the proposed algorithm can asymptotically converge to the correct distribution with a controllable bias under mild conditions. Numerical tests are performed on both synthetic regression problems and learning solutions of elliptic PDE, which demonstrate the accuracy and efficiency of the present work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Bias-Variance Trade-Off in Physics-Informed Neural Networks with Randomized Smoothing for High-Dimensional PDEs

Physics-Informed Neural Networks (PINNs) have triggered a paradigm shift in scientific computing, leveraging mesh-free properties and robust approximation capabilities. While proving effective for low-dimensional partial differential equations (PDEs), the computational cost of PINNs remains a hurdle in high-dimensional scenarios. This is particularly pronounced when computing high-order and high-dimensional derivatives in the physics-informed loss. Randomized Smoothing PINN (RS-PINN) introduces Gaussian noise for stochastic smoothing of the original neural net model, enabling the use of Monte Carlo methods for derivative approximation, which eliminates the need for costly automatic differentiation. Despite its computational efficiency, especially in the approximation of high-dimensional derivatives, RS-PINN introduces biases in both loss and gradients, negatively impacting convergence, especially when coupled with stochastic gradient descent (SGD) algorithms. We present a comprehensive analysis of biases in RS-PINN, attributing them to the nonlinearity of the Mean Squared Error (MSE) loss as well as the intrinsic nonlinearity of the PDE itself. We propose tailored bias correction techniques, delineating their application based on the order of PDE nonlinearity. The derivation of an unbiased RS-PINN allows for a detailed examination of its advantages and disadvantages compared to the biased version. Specifically, the biased version has a lower variance and runs faster than the unbiased version, but it is less accurate due to the bias. To optimize the bias-variance trade-off, we combine the two approaches in a hybrid method that balances the rapid convergence of the biased version with the high accuracy of the unbiased version. In addition to methodological contributions, we present an enhanced implementation of RS-PINN. Extensive experiments on diverse high-dimensional PDEs, including Fokker-Planck, Hamilton-Jacobi-Bellman (HJB), viscous Burgers’, Allen-Cahn, and Sine-Gordon equations, illustrate the bias-variance trade-off and highlight the effectiveness of the hybrid RS-PINN. Empirical guidelines are provided for selecting biased, unbiased, or hybrid versions, depending on the dimensionality and nonlinearity of the specific PDE problem.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

SU-F-J-56: The Connection Between Cherenkov Light Emission and Radiation Absorbed Dose in Proton Irradiated Phantoms

Purpose: Range verification in proton therapy is of great importance. Cherenkov light follows the photon and electron energy deposition in water phantom. The purpose of this study is to investigate the connection between Cherenkov light generation and radiation absorbed dose in a water phantom irradiated with proton beams. Methods: Monte Carlo simulation was performed by employing FLUKA Monte Carlo code to stochastically simulate radiation transport, ionizing radiation dose deposition, and Cherenkov radiation in water phantoms. The simulations were performed for proton beams with energies in the range 50–600 MeV to cover a wide range of proton energies. Results: The mechanism of Cherenkov light production depends on the initial energy of protons. For proton energy with 50–400 MeV energy that is below the threshold (∼483 MeV in water) for Cherenkov light production directly from incident protons, Cherenkov light is produced mainly from the secondary electrons liberated as a result of columbic interactions with the incident protons. For proton beams with energy above 500 MeV, in the initial depth that incident protons have higher energy than the Cherenkov light production threshold, the light has higher intensity. As the slowing down process results in lower energy protons in larger depths in the water phantom, there is a knee point in the Cherenkov light curve vs. depth due to switching the Cherenkov light production mechanism from primary protons to secondary electrons. At the end of the depth dose curve the Cherenkov light intensity does not follow the dose peak because of the lack of high energy protons to produce Cherenkov light either directly or through secondary electrons. Conclusion: In contrast to photon and electron beams, Cherenkov light generation induced by proton beams does not follow the proton energy deposition specially close to the end of the proton range near the Bragg peak.

60 APPLIED LIFE SCIENCES↗
NASA NTRS1994

Stochastic solution to quantum dynamics

The quantum Liouville equation in the Wigner representation is solved numerically by using Monte Carlo methods. For incremental time steps, the propagation is implemented as a classical evolution in phase space modified by a quantum correction. The correction, which is a momentum jump function, is simulated in the quasi-classical approximation via a stochastic process. The technique, which is developed and validated in two- and three- dimensional momentum space, extends an earlier one-dimensional work. Also, by developing a new algorithm, the application to bound state motion in an anharmonic quartic potential shows better agreement with exact solutions in two-dimensional phase space.

John, Sarah↗
DOE OSTI2025

Interpreting experimental measurements of helium bubbles using stochastic cluster dynamics models of heterogeneous nucleation and growth in irradiated ferritic alloys

Among a number of other advantageous features, ferritic/martensitic steels are being considered as fusion reactor structural materials due to their low intrinsic swelling under irradiation. However, under high-energy neutron irradiation, He produced through (n, α) reactions stabilizes vacancy clusters, which then act as seeds for bubble formation and growth, negating the intrinsic swelling resistance of these alloys. Standard models of irradiation damage accumulation and microstructural evolution consider homogeneous nucleation as the basis for bubble formation and growth. However, this generally does not explain the large bubble densities and sizes observed experimentally under a number of different conditions. Here, we propose a more realistic physical model of bubble nucleation, formation, and growth designed to capture recent experimental measurements of He-bubble formation and evolution during co-implantation of He and Fe ions in model ferritic alloys. We find that experimental results are explained only when the following three features are considered simultaneously: (i) heterogeneous nucleation of He-vacancy bubbles at defect sinks (e.g., dislocations, grain boundaries, and second-phase precipitates), (ii) nucleation and growth of bubbles via the ‘trap mutation’ mechanism (i.e., spontaneous production of Frenkel pairs due to absorption of He atoms), and (iii) transition from a growth-limited, He-stabilized bubble regime to a ‘runaway’ void-type growth scenario. The model is implemented into a stochastic cluster dynamics framework that takes advantage of cluster size grouping methods to accelerate the simulations, allowing us to reach 10 dpa of simulated irradiated dose, and to capture cluster sizes in excess of 20 nm. Further, a careful extrapolation exercise conducted assuming classical nucleation theory leads to excellent agreement with the experimental measurements at 50 dpa of irradiation.

36 MATERIALS SCIENCE↗
DOE OSTI2016

Computing the optimal path in stochastic dynamical systems

In stochastic systems, one is often interested in finding the optimal path that maximizes the probability of escape from a metastable state or of switching between metastable states. Even for simple systems, it may be impossible to find an analytic form of the optimal path, and in high-dimensional systems, this is almost always the case. In this article, we formulate a constructive methodology that is used to compute the optimal path numerically. The method utilizes finite-time Lyapunov exponents, statistical selection criteria, and a Newton-based iterative minimizing scheme. The method is applied to four examples. The first example is a two-dimensional system that describes a single population with internal noise. This model has an analytical solution for the optimal path. The numerical solution found using our computational method agrees well with the analytical result. The second example is a more complicated four-dimensional system where our numerical method must be used to find the optimal path. The third example, although a seemingly simple two-dimensional system, demonstrates the success of our method in finding the optimal path where other numerical methods are known to fail. In the fourth example, the optimal path lies in six-dimensional space and demonstrates the power of our method in computing paths in higher-dimensional spaces.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗