Deep learning for presumed probability density function models
Not Available
T. Henry de Frahan, Marc↗
SEARCH · Engineering Papers
Results for “density model”
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.
Modeling of the cold electron plasma density for radiation belt physics
This review focusses strictly on existing plasma density models, including ionospheric source models, empirical density models, physics-based and machine-learning density models. This review is framed in the context of radiation belt physics and space weather codes. The review is limited to the most commonly used models or to models recently developed and promising. A great variety of conditions is considered such as the magnetic local time variation, geomagnetic conditions, ionospheric source regions, radial and latitudinal dependence, and collisional vs. collisionless conditions. These models can serve to complement satellite observations of the electron plasma density when data are lacking, are for most of them commonly used in radiation belt physics simulations and can improve our understanding of the plasmasphere dynamics.
79 ASTRONOMY AND ASTROPHYSICS↗
Chemical insight from density functional modeling of molecular adsorption: Tracking the bonding and diffusion of anthracene derivatives on Cu(111) with molecular orbitals
We present a method of analyzing the results of density functional modeling of molecular adsorption in terms of an analogue of molecular orbitals. This approach permits intuitive chemical insight into the adsorption process. Applied to a set of anthracene derivates (anthracene, 9,10-anthraquinone, 9,10-dithioanthracene, and 9,10-diselenonanthracene), we follow the electronic states of the molecules that are involved in the bonding process and correlate them to both the molecular adsorption geometry and the species’ diffusive behavior. We additionally provide computational code to easily repeat this analysis on any system.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Uniform modelling of the stellar density of thirteen tidal streams within the Galactic halo
ABSTRACT We present the results of fitting a flexible stellar stream density model to a collection of thirteen streams around the Milky Way, using photometric data from DES, DECaLS, and Pan-STARRS. We construct density maps for each stream and characterize their tracks on the sky, width, and distance modulus curves along the length of each stream. We use these measurements to compute lengths and total luminosities of streams and identify substructures. Several streams show prominent substructures, such as stream broadening, gaps, large deviations of stream tracks, and sharp changes in stream densities. Examining the group of streams as a population, as expected we find that streams with globular cluster progenitors are typically narrower than those with dwarf galaxy progenitors, with streams around 100 pc wide showing overlap between the two populations. We also note the average luminosity of globular cluster streams is significantly lower than the typical luminosity of intact globular clusters. The likely explanation is that observed globular cluster streams preferentially come from lower luminosity and lower density clusters. The stream measurements done in a uniform manner presented here will be helpful for more detailed stream studies such as identifying candidate stream members for spectroscopic follow up and stellar stream dynamical modelling.
79 ASTRONOMY AND ASTROPHYSICS↗
Projected density matrix embedding theory with applications to the two-dimensional Hubbard model
Density matrix embedding theory (DMET) is a quantum embedding theory for strongly correlated systems. From a computational perspective, one bottleneck in DMET is the optimization of the correlation potential to achieve self-consistency, especially for heterogeneous systems of large size. We introduce a new approach, called projected DMET (p-DMET), which achieves self-consistency without needing to optimize the correlation potential. We demonstrate the performance of p-DMET on the two-dimensional Hubbard model.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Diffusion-Model-Assisted Supervised Learning of Generative Models for Density Estimation
Here, we present a supervised learning framework of training generative models for density estimation. Generative models, including generative adversarial networks (GANs), normalizing flows, and variational auto-encoders (VAEs), are usually considered as unsupervised learning models, because labeled data are usually unavailable for training. Despite the success of the generative models, there are several issues with the unsupervised training, e.g., requirement of reversible architectures, vanishing gradients, and training instability. To enable supervised learning in generative models, we utilize the score-based diffusion model to generate labeled data. Unlike existing diffusion models that train neural networks to learn the score function, we develop a training-free score estimation method. This approach uses mini-batch-based Monte Carlo estimators to directly approximate the score function at any spatial-temporal location in solving an ordinary differential equation (ODE), corresponding to the reverse-time stochastic differential equation (SDE). This approach can offer both high accuracy and substantial time savings in neural network training. Once the labeled data are generated, we can train a simple, fully connected neural network to learn the generative model in the supervised manner. Compared with existing normalizing flow models, our method does not require the use of reversible neural networks and avoids the computation of the Jacobian matrix. Compared with existing diffusion models, our method does not need to solve the reverse-time SDE to generate new samples. As a result, the sampling efficiency is significantly improved. We demonstrate the performance of our method by applying it to a set of 2D datasets as well as real data from the University of California Irvine (UCI) repository.
97 MATHEMATICS AND COMPUTING↗
Variable Density Flow Modeling for Simulation Framework for Regional Geologic CO{sub 2} Storage Along Arches Province of Midwestern United States
The Arches Province in the Midwestern U.S. has been identified as a major area for carbon dioxide (CO{sub 2}) storage applications because of the intersection of Mt. Simon sandstone reservoir thickness and permeability. To better understand large-scale CO{sub 2} storage infrastructure requirements in the Arches Province, variable density scoping level modeling was completed. Three main tasks were completed for the variable density modeling: Single-phase, variable density groundwater flow modeling; Scoping level multi-phase simulations; and Preliminary basin-scale multi-phase simulations. The variable density modeling task was successful in evaluating appropriate input data for the Arches Province numerical simulations. Data from the geocellular model developed earlier in the project were translated into preliminary numerical models. These models were calibrated to observed conditions in the Mt. Simon, suggesting a suitable geologic depiction of the system. The initial models were used to assess boundary conditions, calibrate to reservoir conditions, examine grid dimensions, evaluate upscaling items, and develop regional storage field scenarios. The task also provided practical information on items related to CO{sub 2} storage applications in the Arches Province such as pressure buildup estimates, well spacing limitations, and injection field arrangements. The Arches Simulation project is a three-year effort and part of the United States Department of Energy (U.S. DOE)/National Energy Technology Laboratory (NETL) program on innovative and advanced technologies and protocols for monitoring/verification/accounting (MVA), simulation, and risk assessment of CO{sub 2} sequestration in geologic formations. The overall objective of the project is to develop a simulation framework for regional geologic CO{sub 2} storage infrastructure along the Arches Province of the Midwestern U.S.
54 ENVIRONMENTAL SCIENCES↗
Integrating land surface phenology with cluster density and size improves spatially explicit models of animal density
Not Available
Butler, Matthew J.↗
SU-E-T-236: Monte Carlo Calculations for Radiosurgery of the Clivus
Purpose: We sought to determine if dosimetric differences exist between Ray-Trace (RT) and Monte-Carlo (MC) calculation algorithms for radiosurgery for tumors of the clivus, given their proximity to air-tissue inhomogeneity. Methods: We retrospectively identified 10 patients with a tumor of the clivus where the target was located near an air-tissue interface. CyberKnife treatment was delivered in 1 to 5 fractions. Plans originally calculated with the RT algorithm were re-calculated with the same monitor units (MU) with the MC algorithm. Similarly, plans originally calculated with MC were recalculated with RT. MC calculations were performed using High Resolution with 2% uncertainty on the MultiPlan 4.6 planning system. If the RT plans were originally calculated using water-air density model, they were re-calculated with Body standard density model, and then compared with MC plans computed with Body standard density model. The maximum dose (Dmax) to the planning target volumes (PTV) and critical structures, minimum dose (Dmin) and coverage percentage for the PTV were calculated. Results: For 8 cases, MC PTV Dmin was a median 5% (range 0–10%) and Dmax a median 6% (range 3–9%) higher than for RT. For 2 cases, MC PTV Dmin was 6 and 2% and Dmax was 1 and 3% lower than for RT. MC PTV Dmin and Dmax maybe higher or lower than for RT even though the target coverage for MC was less than for RT (median 4%, range 0–13%) with the same MU used. The degree of differences depends on the location of the target relative to the air-tissue inhomogeneity. Examples are shown that the targets are underdosed near the air-tissue inhomogenity. Conclusion: Monte Carlo dose calculations are recommended for targets near tissue homogeneity such as the clivus. Assuming that Monte Carlo more closely approximates the true dosimetry, the use of Ray Trace could overestimate target coverage.
60 APPLIED LIFE SCIENCES↗
Multiplicative noise and the diffusion of conserved densities
Stochastic fluid dynamics governs the long time tails of hydrodynamic correlation functions, and the critical slowing down of relaxation phenomena in the vicinity of a critical point in the phase diagram. In this work we study the role of multiplicative noise in stochastic fluid dynamics. Multiplicative noise arises from the dependence of transport coefficients, such as the diffusion constants for charge and momentum, on fluctuating hydrodynamic variables. We study long time tails and relaxation in the diffusion of a conserved density (model B), and a conserved density coupled to the transverse momentum density (model H). Careful attention is paid to fluctuation-dissipation relations. We observe that multiplicative noise contributes at the same order as non-linear interactions in model B, but is a higher order correction to the relaxation of a scalar density and the tail of the stress tensor correlation function in model H.
73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
Density-Matrix Model for Photon-Driven Transport in Quantum Cascade Lasers
In this report we develop a time-dependent density-matrix model to study photon-assisted (PA) electron transport in quantum cascade lasers. The Markovian equation of motion for the density matrix in the presence of an optical field is solved for an arbitrary field amplitude. Level-broadening terms emerge from microscopic Hamiltonians and supplant the need for empirical parameters that are often employed in related approaches. We show that, in quantum cascade lasers with diagonal design, photon resonances have a pronounced impact on electron dynamics around and above the lasing threshold, an effect that stems from the large spatial separation between the upper and lower lasing states. With the inclusion of PA tunneling, the calculated current density and output power are in good agreement with experiment.
36 MATERIALS SCIENCE↗
DeePKS-kit: A package for developing machine learning-based chemically accurate energy and density functional models
We introduce DeePKS-kit, an open-source software package for developing machine learning based energy and density functional models. DeePKS-kit is interfaced with PyTorch, an open-source machine learning library, and PySCF, an ab initio computational chemistry program that provides simple and customized tools for developing quantum chemistry codes. It supports the DeePHF and DeePKS methods. In addition to explaining the details in the methodology and the software, we also provide an example of developing a chemically accurate model for water clusters.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Progress on Complex Langevin simulations of a finite density matrix model for QCD
We study the Stephanov model, which is an RMT model for QCD at finite density, using the Complex Langevin algorithm. Naive implementation of the algorithm shows convergence towards the phase quenched or quenched theory rather than to intended theory with dynamical quarks. A detailed analysis of this issue and a potential resolution of the failure of this algorithm are discussed. We study the effect of gauge cooling on the Dirac eigenvalue distribution and time evolution of the norm for various cooling norms, which were specifically designed to remove the pathologies of the complex Langevin evolution. The cooling is further supplemented with a shifted representation for the random matrices. Unfortunately, none of these modifications generate a substantial improvement on the complex Langevin evolution and the final results still do not agree with the analytical predictions.
Bloch, Jacques↗
Diamonds in the rough: a strong case for the inclusion of weak-intensity X-ray diffraction data
Here, new evidence is provided to show that the inclusion of weak-intensity, high-resolution X-ray diffraction data helps to improve the quality of experimental phases by imposing proper constraints on electron-density models during noncrystallographic symmetry averaging. Overwhelming evidence exists to show that the inclusion of weak-intensity, high-resolution X-ray diffraction data helps improve the refinement of atomic models by imposing strong constraints on individual and overall temperature B factors and thus the quality of crystal structures. Some researchers consider these data to be of little value and opt to discard them during data processing, particularly at medium and low resolution, at which individual B factors of atomic models cannot be refined. Here, new evidence is provided to show that the inclusion of these data helps to improve the quality of experimental phases by imposing proper constraints on electron-density models during noncrystallographic symmetry (NCS) averaging. Using electron-density correlation coefficients as criteria, the resolution of data has successfully been extended from 3.1 to 2.5 Å resolution with redundancy-independent merging R factors from below 100% to about 310%. It is further demonstrated that phase information can be fully extracted from observed amplitudes through de novo NCS averaging. Averaging starts with uniform density inside double-shelled spherical masks and NCS matrices that are derived from bound heavy-atom clusters at the vertices of cuboctahedrally symmetric protein particles.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
Polarizable Water Potential Derived from a Model Electron Density
A new empirical potential for efficient, large scale molecular dynamics simulation of water is presented. The HIPPO (Hydrogen-like Intermolecular Polarizable POtential) force field is based upon the model electron density of a hydrogen-like atom. This framework is used to derive and parametrize individual terms describing charge penetration damped permanent electrostatics, damped polarization, charge transfer, anisotropic Pauli repulsion, and damped dispersion interactions. Initial parameter values were fit to Symmetry Adapted Perturbation Theory (SAPT) energy components for ten water dimer configurations, as well as the radial and angular dependence of the canonical dimer. The SAPT-based parameters were then systematically refined to extend the treatment to water bulk phases. The final HIPPO water model provides a balanced representation of a wide variety of properties of gas phase clusters, liquid water, and ice polymorphs, across a range of temperatures and pressures. This water potential yields a rationalization of water structure, dynamics, and thermodynamics explicitly correlated with an ab initio energy decomposition, while providing a level of accuracy comparable or superior to previous polarizable atomic multipole force fields. The HIPPO water model serves as a cornerstone around which similarly detailed physics-based models can be developed for additional molecular species.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Dislocation Density-Based Model for Flow Behavior of a Near-α Titanium Alloy Considering Effects of Initial Lamellar Thickness
A dislocation density-based constitutive model that considered the effects of initial lamellar thickness was proposed to predict the flow behavior of a near-α high-temperature Ti-5.4Al-3.7Sn-3.3Zr-0.5Mo-0.4Si alloy. The evolution equations of dislocation density were improved by introducing the influence of the dynamic globularization of lamellar microstructure with different initial thicknesses. The evolution equations and Hall-Petch behavior were coupled into the constitutive model, which was composed of athermal and thermal stresses. According to the established constitutive equations, the variation and contributions of corresponding stress components were analyzed. The calculated dislocation densities were compared with the experimental values obtained by electron backscattered diffraction data to verify the model. Moreover, the flow stresses of the studied alloy with different initial lamellar microstructures under various deformation conditions were predicted. Good agreements between the predicted and experimental results were obtained. Additionally, steady-state flow behavior was further analyzed and the relationship between the calculated steady-state flow stress and the equiaxed α-phase size was quantitatively obtained. The results indicated the validity of the proposed constitutive model for predicting the flow behavior of the alloy with different initial lamellar microstructures over a range of deformation temperatures and strain rates.
36 MATERIALS SCIENCE↗
A TYPE II RADIO BURST WITHOUT A CORONAL MASS EJECTION
Type II radio bursts are thought to be a signature of coronal shocks. In this paper, we analyze a short-lived type II burst that started at 07:40 UT on 2011 February 28. By carefully checking white-light images, we find that the type II radio burst is not accompanied by a coronal mass ejection, only by a C2.4 class flare and narrow jet. However, in the EUV images provided by the Atmospheric Imaging Assembly on board the Solar Dynamics Observatory, we find a wave-like structure that propagated at a speed of ∼600 km s{sup −1} during the burst. The relationship between the type II radio burst and the wave-like structure is, in particular, explored. For this purpose, we first derive the density distribution under the wave by the differential emission measure method, which is used to restrict the empirical density model. We then use the restricted density model to invert the speed of the shock that produces the observed frequency drift rate in the dynamic spectrum. The inverted shock speed is similar to the speed of the wave-like structure. This implies that the wave-like structure is most likely a coronal shock that produces the type II radio burst. We also examine the evolution of the magnetic field in the flare-associated active region and find continuous flux emergence and cancellation taking place near the flare site. Based on these facts, we propose a new mechanism for the formation of the type II radio burst, i.e., the expansion of the strongly inclined magnetic loops after reconnecting with a nearby emerging flux acts as a piston to generate the shock wave.
79 ASTRONOMY AND ASTROPHYSICS↗