Search NASA⌕ Search

SEARCH · Search NASA

Results for “Stochastic optimization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,405 records · Page 78

Uncertainty quantification in scientific machine learning: Methods, metrics, and comparisons

Neural networks (NNs) are currently changing the computational paradigm on how to combine data with mathematical laws in physics and engineering in a profound way, tackling challenging inverse and ill-posed problems not solvable with traditional methods. However, quantifying errors and uncertainties in NN-based inference is more complicated than in traditional methods. This is because in addition to aleatoric uncertainty associated with noisy data, there is also uncertainty due to limited data, but also due to NN hyperparameters, overparametrization, optimization and sampling errors as well as model misspecification. Although there are some recent works on uncertainty quantification (UQ) in NNs, there is no systematic investigation of suitable methods towards quantifying the total uncertainty effectively and efficiently even for function approximation, and there is even less work on solving partial differential equations and learning operator mappings between infinite-dimensional function spaces using NNs. In this work, we present a comprehensive framework that includes uncertainty modeling, new and existing solution methods, as well as evaluation metrics and post-hoc improvement approaches. Further, to demonstrate the applicability and reliability of our framework, we present an extensive comparative study in which various methods are tested on prototype problems, including problems with mixed input-output data, and stochastic problems in high dimensions. In the Appendix, we include a comprehensive description of all the UQ methods employed. Further, to help facilitate the deployment of UQ in Scientific Machine Learning research and practice, we present and develop in [1] an open-source Python library (github.com/Crunch-UQ4MI/neuraluq), termed NeuralUQ, that is accompanied by an educational tutorial and additional computational experiments.

11 physics-informed neural networks↗

Optimal time schedule for adiabatic evolution

We show that, provided dephasing is taken into account, there is a unique timetable which maximizes the fidelity with a target state in adiabatic evolutions. The optimum has constant tunneling rate along the path. Application to quantum search algorithms recovers the Grover result for appropriate scaling of the dephasing with the size of the database. Moreover, the Grover bound imposes constraints on the dephasing rates of systems coupled to a universal Markovian bath.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The Stability of Fiber Spectrographs in the Faint-source Regime

The use of optical fibers in astronomical instrumentation offers high-multiplex and light-gathering flexibility. However, with most previous fiber spectrographs optimized for large fields of view on modest-aperture telescopes, the performance of fibers in the context of faint targets on large telescopes remains largely untested. In this paper, we evaluate aspects of fiber stability, especially as they apply in the context of precision sky subtraction of faint sources at modest spectral resolution ($R$ ~ 3000). After introducing a framework for describing potential systematic errors, we use publicly available data from existing instruments, including instrumentation used by the fourth-generation Sloan Digital Sky Survey’s MaNGA project (MaNGA: Mapping Nearby Galaxies at Apache Point Observatory) and the Very Large Telescope’s FLAMES: Fiber Large Array Multi Element Spectrograph. We isolate sources of fiber systematics and estimate the observed amplitude of persistent residuals as well as stochastic noise contributions resulting from changing fiber stresses. Comparing these levels against their impact on various sky subtraction schemes demonstrates that 0.1% precision sky subtraction with fiber instruments is possible. As a demonstration, we show that the MaNGA instrument can deliver 0.2% residuals on bright near-IR sky lines with nonlocal sky subtraction, if pseudo-slit limitations are addressed by allocating 50% of its fibers to sky. We further highlight recently published deep exposures that achieved a 1$σ$ background level of 27.6 AB per square arc second, equivalent to a precision of 0.2% of the sky background continuum.

79 ASTRONOMY AND ASTROPHYSICS↗

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↗

Capacitated arc routing problem and its extensions in waste collection

Capacitated arc routing problem (CARP) is the youngest generation of graph theory that focuses on solving the edge/arc routing for optimality. Since many years, operational research devoted to CARP counterpart, known as vehicle routing problem (VRP), which does not fit to several real cases such like waste collection problem and road maintenance. In this paper, we highlighted several extensions of capacitated arc routing problem (CARP) that represents the real-life problem of vehicle operation in waste collection. By purpose, CARP is designed to find a set of routes for vehicles that satisfies all pre-setting constraints in such that all vehicles must start and end at a depot, service a set of demands on edges (or arcs) exactly once without exceeding the capacity, thus the total fleet cost is minimized. We also addressed the differentiation between CARP and VRP in waste collection. Several issues have been discussed including stochastic demands and time window problems in order to show the complexity and importance of CARP in the related industry. A mathematical model of CARP and its new version is presented by considering several factors such like delivery cost, lateness penalty and delivery time.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Surrogate modeling of Monte Carlo radiation transport with convolutional neural networks for shielding optimization

Here, we present a machine learning (ML)-based surrogate model using convolutional neural networks (CNN) designed to emulate the attenuation of neutron fields as they pass through various shielding materials. This model can compute the outgoing neutron flux almost instantaneously and achieves reasonable accuracy compared to traditional Monte Carlo (MC)-based codes, which are computationally intensive. This emulator alleviates the complexity of neutron radiation transport through shielding materials by reducing the dimensionality and enables shielding optimization for a known radiation environment. This optimization process, which would have taken an unrealistic timeline due to several complex radiation transport simulations, can now be achieved in minutes, thus increasing computational capabilities in radiation shielding assessment. We demonstrate the applications of this emulator in computing effective dose rates and optimizing shielding solutions for a heavy-ion accelerator facility, such as the Facility for Rare Isotope Beams, where secondary neutrons produced via beam interactions dominate the radiation environment.

accelerator shielding↗

TH-A-BRD-01: Radiation Biology for Radiation Therapy Physicists

Mechanisms by which radiation kills cells and ways cell damage can be repaired will be reviewed. The radiobiological parameters of dose, fractionation, delivery time, dose rate, and LET will be discussed. The linear-quadratic model for cell survival for high and low dose rate treatments and the effect of repopulation will be presented and discussed. The rationale for various radiotherapy techniques such as conventional fractionation, hyperfractionation, hypofractionation, and low and high dose rate brachytherapy, including permanent implants, will be presented. The radiobiological principles underlying radiation protection guidelines and the different radiation dosimetry terms used in radiation biology and in radiation protection will be reviewed. Human data on radiation induced cancer, including increases in the risk of second cancers following radiation therapy, as well as data on radiation induced tissue reactions, such as cardiovascular effects, for follow up times up to 20–40 years, published by ICRP, NCRP and BEIR Committees, will be examined. The latest risk estimates per unit dose will be presented. Their adoption in recent radiation protection standards and guidelines and their impact on patient and workers safety in radiotherapy will be discussed. Biologically-guided radiotherapy (BGRT) provides a systematic method to derive prescription doses that integrate patient-specific information about tumor and normal tissue biology. Treatment individualization based on patient-specific biology requires the identification of biological objective functions to facilitate the design and comparison of competing treatment modalities. Biological objectives provide a more direct approach to plan optimization instead of relying solely on dose-based surrogates and can incorporate factors that alter radiation response, such as DNA repair, tumor hypoxia, and relative biological effectiveness. We review concepts motivating biological objectives and provide examples of how they might be used to address clinically relevant problems. Underlying assumptions and limitations of existing models and their proper application will be discussed. This multidisciplinary educational session combines the fundamentals of radiobiology for radiation therapy and radiation protection with the practical application of biophysical models for treatment planning and evaluation. Learning Objectives: To understand fractionation in teletherapy and dose rate techniques in brachytherapy. To understand how the linear-quadratic models the effect of radiobiological parameters for radiotherapy. To understand the radiobiological basis of radiation protection standards applied to radiotherapy. To distinguish between stochastic effects and tissue reactions. To learn how to apply concepts of biological effective dose and RBE-weighted dose and to incorporate biological factors that alter radiation response. To discuss clinical strategies to increase therapeutic ratio, i.e., maximize local control while minimizing the risk of acute and late normal tissue effects.

60 APPLIED LIFE SCIENCES↗

The root cause of disruptive NTMs and paths to stable operation in DIII-D ITER baseline scenario plasmas

Analyses of the DIII-D ITER Baseline Scenario database support that the disruptive m,n=2,1 magnetic islands are pressure gradient driven, non-linear instabilities seeded in a sequence of stochastic transient magnetic perturbations, and that the current profile relaxation does not affect the m,n=2,1 island onset rate. At low torque, these Neoclassical Tearing Modes are most commonly seeded by non-linear 3-wave coupling when the differential rotation between the q=1 & q=2 rational surfaces approaches zero. Lack of statistically significant difference between the current profiles of stable and unstable states, as well as lack of correlation between the tearing mode onset rate and the current profile relaxation both reject causality between the current profile evolution and the 2,1 magnetic island onsets in these plasmas. These support that preserving the differential rotation between the q=1 and q=2 rational surfaces is key to long pulse stable operation in the plasma scenario planned for ITER, while optimization of the current profile within the explored parameter space may lead to much weaker improvements than sustaining the differential rotation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Growth Optimal Portfolio Selection Under Proportional Transaction Costs with Obligatory Diversification

A continuous time long run growth optimal or optimal logarithmic utility portfolio with proportional transaction costs consisting of a fixed proportional cost and a cost proportional to the volume of transaction is considered. The asset prices are modeled as exponent of diffusion with jumps whose parameters depend on a finite state Markov process of economic factors. An obligatory portfolio diversification is introduced, accordingly to which it is required to invest at least a fixed small portion of our wealth in each asset.

97 MATHEMATICS AND COMPUTING↗

An automated approach to the alignment of compound refractive lenses

Compound refractive lenses (CRLs) are established X-ray focusing optics, and are used to focus the beam or image the sample in many beamlines at X-ray facilities. While CRLs are quite established, the stack of single lens elements affords a very small numerical aperture because of the thick lens profile, making them far more difficult to align than classical optical lenses that obey the thin-lens approximation. This means that the alignment must be very precise and is highly sensitive to changes to the incident beam, often requiring regular readjustments. Some groups circumvent the full realignment procedure by using engineering controls ( e.g. mounting optics) that sacrifice some of the beam's focusing precision, i.e. spot size, or resolution. While these choices minimize setup time, there are clear disadvantages. This work presents a new automated approach to align CRLs using a simple alignment apparatus that is easy to adapt and install at different types of X-ray experiments or facilities. This approach builds on recent CRL modeling efforts, using an approach based on the Stochastic Nelder–Mead (SNM) simplex method. This method is outlined and its efficacy is demonstrated with numerical simulation that is tested in real experiments conducted at the Advanced Photon Source to confirm its performance with a synchrotron beam. This work provides an opportunity to automate key instrumentation at X-ray facilities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Prediction of gas hydrate saturation using machine learning and optimal set of well-logs

We report resistivity and acoustic logs are widely used to estimate gas hydrate saturation in various sedimentary systems using one of the two popular methods ((1) acoustic velocity and (2) electrical resistivity), but the limitations of these two methods are often overlooked, which include (i) well-specific calibration of empirical exponents in the electrical resistivity method, (ii) assumption of known pore morphology for gas hydrates in the acoustic velocity method, and (iii) presence of unknown mineralogy and bulk modulus terms in the acoustic velocity method. NMR-density porosity-derived gas hydrate saturation based on the analysis of the transverse magnetization relaxation time (T2) is considered the most precise method, but acquisition of NMR-based logs is limited at relatively recent drilled sites; additionally, its use in conventional oil and gas reservoirs is not that common due to higher cost and operational deployment limitations associated with acquiring NMR well-logs. This study proposes a new method that predicts gas hydrate saturation (S h ) for any well using porosity, bulk density, and compressional wave (P wave) velocity well-logs with neural network (or stochastic gradient descent regression) without any well-specific calibration and/or other aforementioned shortcomings of the existing methods. The method is developed by examining the underlying dependency between S h and different combinations of well-logs, chosen from 6 routine logs, with 12 different machine learning (ML) algorithms. The accuracy of the proposed method in predicting S h is ~ 84%, which is better than the accuracy of seismic and electrical resistivity methods (≤ 75%) per the results reported by three different studies. The robustness of the method in the specific case of permafrost-associated gas hydrates is demonstrated with well-log data from two wells drilled on the Alaska North Slope.

58 GEOSCIENCES↗

Function Preserving Projection

With function preserving projection (FPP), we aim to address high-dimensional data exploration challenges common in many scientific applications. Instead of preserving the global or local geometric structure of a dataset like traditional dimensionality reduction methods (e.g., principal component analysis, t-stochastic neighborhood embedding), the proposed approach finds a linear projection that reveal interpretable (yet potentially nonlinear) pattern of the properties of interests defined on a highdimensional domain (e.g., simulation yield defined on the input parameter space). The key insight driving FPP is the fact that while non-linear embeddings are difficult to interpret, humans are highly skilled in understanding complex visual patterns. FPP lets users select the complexity of a pattern they consider interpretable, allows to find a joint projection that explains multiple functions, and provides a confidence statistic (akin to a p-value) indicating how salient the resulting pattern is. We demonstrate the efficacy of the proposed method on several synthetic and scientific datasets for both continuous (regression) and discrete (classification) problems of interests and illustrate the scalability of FPP by finding an optimal projection for close to 40 gigabytes in around 10 minutes.

Liu, SHUSEN↗

A phase transition for finding needles in nonlinear haystacks with LASSO artificial neural networks

To fit sparse linear associations, a LASSO sparsity inducing penalty with a single hyperparameter provably allows to recover the important features (needles) with high probability in certain regimes even if the sample size is smaller than the dimension of the input vector (haystack). More recently learners known as artificial neural networks (ANN) have shown great successes in many machine learning tasks, in particular fitting nonlinear associations. Small learning rate, stochastic gradient descent algorithm and large training set help to cope with the explosion in the number of parameters present in deep neural networks. Yet few ANN learners have been developed and studied to find needles in nonlinear haystacks. Driven by a single hyperparameter, our ANN learner, like for sparse linear associations, exhibits a phase transition in the probability of retrieving the needles, which we do not observe with other ANN learners. To select our penalty parameter, we generalize the universal threshold of Donoho and Johnstone (Biometrika 81(3):425–455, 1994) which is a better rule than the conservative (too many false detections) and expensive cross-validation. In the spirit of simulated annealing, we propose a warm-start sparsity inducing algorithm to solve the high-dimensional, non-convex and non-differentiable optimization problem. We perform simulated and real data Monte Carlo experiments to quantify the effectiveness of our approach.

97 MATHEMATICS AND COMPUTING↗

Progress Summary: Simulating stochastic and additively manufactured foams at large strains and high strain rates

Motivation: Understanding the mechanical response of polymer foams is essential in developing system material models and in designing replacements for legacy pads and cushions. Simulations can help elucidate structure-property relationships and compare local stress distributions to understand aging behavior and improve designs. In many applications,these materials experience significant loadings and computations need the capability to reproduce those conditions. Significance: In designing the compressive and shear response of additively manufactured foams, we expect the curvature of the substrate to influence their mechanical behavior. Progress:We have the capability of mapping planar structures to arbitrary curved surfaces and are currently developing the software and procedures to translate the mapped pattern into a geometrical model that can be imported to either finite element or material point software to evaluate the mechanical response. Figure 1 depicts a four-layer face-centered tetragonal pattern (FCT) mapped to a hemispherical surface. Because the mapping cannot exactly preserve both distances and angles, we expect some local variation in mechanical properties and porosity. These models will help determine the sensitivity of these variations to structure and curvature, optimize the mapping procedure, and inform the design of foams for non-planar surfaces.

36 MATERIALS SCIENCE↗

Capacity of a bosonic memory channel with Gauss-Markov noise

We address the classical capacity of a quantum bosonic memory channel with additive noise, subject to an input energy constraint. The memory is modeled by correlated noise emerging from a Gauss-Markov process. Under reasonable assumptions, we show that the optimal modulation results from a 'quantum water-filling' solution above a certain input energy threshold, similar to the optimal modulation for parallel classical Gaussian channels. We also derive analytically the optimal multimode input state above this threshold, which enables us to compute the capacity of this memory channel in the limit of an infinite number of modes. The method can also be applied to a more general noise environment which is constructed by a stationary Gauss process. The extension of our results to the case of broadband bosonic channels with colored Gaussian noise should also be straightforward.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Final Technical Report: Distributed Controls for High Penetrations of Renewables

The goal of this effort was to apply four potential control analysis/design approaches to the design of distributed grid control systems to address the impact of latency and communications uncertainty with high penetrations of photovoltaic (PV) generation. The four techniques considered were: optimal fixed structure control; Nyquist stability criterion; vector Lyapunov analysis; and Hamiltonian design methods. A reduced order model of the Western Electricity Coordinating Council (WECC) developed for the Matlab Power Systems Toolbox (PST) was employed for the study, as well as representative smaller systems (e.g., a two-area, three-area, and four-area power system). Excellent results were obtained with the optimal fixed structure approach, and the methodology we developed was published in a journal article. This approach is promising because it offers a method for designing optimal control systems with the feedback signals available from Phasor Measurement Unit (PMU) data as opposed to full state feedback or the design of an observer. The Nyquist approach inherently handles time delay and incorporates performance guarantees (e.g., gain and phase margin). We developed a technique that works for moderate sized systems, but the approach does not scale well to extremely large system because of computational complexity. The vector Lyapunov approach was applied to a two area model to demonstrate the utility for modeling communications uncertainty. Application to large power systems requires a method to automatically expand/contract the state space and partition the system so that communications uncertainty can be considered. The Hamiltonian Surface Shaping and Power Flow Control (HSSPFC) design methodology was selected to investigate grid systems for energy storage requirements to support high penetration of variable or stochastic generation (such as wind and PV) and loads. This method was applied to several small system models.

14 SOLAR ENERGY↗

Towards Improved Understanding of the Applicability of Uncertainty Forecasts in the Electric Power Industry

Around the world wind energy is starting to become a major energy provider in electricity markets, as well as participating in ancillary services markets to help maintain grid stability. The reliability of system operations and smooth integration of wind energy into electricity markets has been strongly supported by years of improvement in weather and wind power forecasting systems. Deterministic forecasts are still predominant in utility practice although truly optimal decisions and risk hedging are only possible with the adoption of uncertainty forecasts. One of the main barriers for the industrial adoption of uncertainty forecasts is the lack of understanding of its information content (e.g., its physical and statistical modeling) and standardization of uncertainty forecast products, which frequently leads to mistrust towards uncertainty forecasts and their applicability in practice. Our paper aims at improving this understanding by establishing a common terminology and reviewing the methods to determine, estimate, and communicate the uncertainty in weather and wind power forecasts. This conceptual analysis of the state of the art highlights that: (i) end-users should start to look at the forecast's properties in order to map different uncertainty representations to specific wind energy-related user requirements; (ii) a multidisciplinary team is required to foster the integration of stochastic methods in the industry sector. Furthermore, a set of recommendations for standardization and improved training of operators are provided along with examples of best practices.

24 POWER TRANSMISSION AND DISTRIBUTION↗