Search NASA⌕ Search

SEARCH · Search NASA

Results for “gradient 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 181 records · Page 10

Poster — Thur Eve — 03: Application of the non-negative matrix factorization technique to [{sup 11}C]-DTBZ dynamic PET data for the early detection of Parkinson's disease

In this work, a class of non-negative matrix factorization (NMF) technique known as alternating non-negative least squares, combined with the projected gradient method, is used to analyze twenty-five [{sup 11}C]-DTBZ dynamic PET/CT brain data. For each subject, a two-factor model is assumed and two factors representing the striatum (factor 1) and the non-striatum (factor 2) tissues are extracted using the proposed NMF technique and commercially available factor analysis software “Pixies”. The extracted factor 1 and 2 curves represent the binding site of the radiotracer and describe the uptake and clearance of the radiotracer by soft tissues in the brain, respectively. The proposed NMF technique uses prior information about the dynamic data to obtain sample time-activity curves representing the striatum and the non-striatum tissues. These curves are then used for “warm” starting the optimization. Factor solutions from the two methods are compared graphically and quantitatively. In healthy subjects, radiotracer uptake by factors 1 and 2 are approximately 35–40% and 60–65%, respectively. The solutions are also used to develop a factor-based metric for the detection of early, untreated Parkinson's disease. The metric stratifies healthy subjects from suspected Parkinson's patients (based on the graphical method). The analysis shows that both techniques produce comparable results with similar computational time. The “semi-automatic” approach used by the NMF technique allows clinicians to manually set a starting condition for “warm” starting the optimization in order to facilitate control and efficient interaction with the data.

60 APPLIED LIFE SCIENCES↗

Fast and Scalable FFT-Based GPU-Accelerated Algorithms for Block-Triangular Toeplitz Matrices with Application to Linear Inverse Problems Governed by Autonomous Dynamical Systems

In this work, we present an efficient and scalable algorithm for performing matrix-vector multiplications (matvecs) for block Toeplitz matrices. Such matrices, which are shift-invariant with respect to their blocks, arise in the context of solving inverse problems governed by autonomous systems, and time-invariant systems in particular. In this article, we consider inverse problems that infer unknown parameters from observational data of a linear time-invariant dynamical system given in the form of partial differential equations (PDEs). Matrix-free Newton-conjugate-gradient methods are often the gold standard for solving these inverse problems, but they require numerous actions of the Hessian on a vector. Matrix-free adjoint-based Hessian matvecs require solution of a pair of linearized forward/adjoint PDE solves per Hessian action, which may be prohibitive for large-scale inverse problems. Time invariance of the forward PDE problem leads to a block Toeplitz structure of the discretized parameter-to-observable (p2o) map defining the mapping from inputs (parameters) to outputs (observables) of the PDEs. This block Toeplitz structure enables us to exploit two key properties: (1) compact storage of the p2o map and its adjoint, and (2) efficient fast Fourier transform–based Hessian matvecs. The proposed algorithm is mapped onto large multi-GPU clusters and achieves more than 80% of peak bandwidth on NVIDIA A100 GPUs. Excellent weak scaling is shown for up to 48 A100 GPUs. For the targeted problems, the implementation executes Hessian matvecs within fractions of a second, which is orders of magnitude faster than can be achieved by conventional matrix-free Hessian matvecs via forward/adjoint PDE solves.

97 MATHEMATICS AND COMPUTING↗

Optipack

Used to solve basic unconstrained optimization problems that can produce smooth accurate gradients. These are standard published unconstrained optimization such as are described in the book Nocedal and Wright Numerical Optimization: 2nd edition, 2006. This software is generic to any mathematical optimization problem where smooth derivatives exist. There is no specific connection or mention whatsoever to any specific application, period. You cannot find more general mathematical software.OptiPack contains some simple optimization algorithms built on Thyra. Currently, only a few different nonlinear conjugate gradient methods are implemented.

Bartlett, Roscoe↗

Deep vadose zone respiration contributions to CO2 fluxes from a semi-arid floodplain, Vadose Zone Journal: Dataset

This data package provides lists of parameter values presented in the paper “Deep vadose zone respiration contributions to CO2 fluxes from a semi-arid floodplain” by Tokunaga et al., Vadose Zone J. 2016. Each table or figure from that paper that contains numerical information is represented as a sheet within the Excel file. Table 1 lists depth profile values for particle-size, bulk density, and parameters for correlating moisture contents with matric potentials. Table 2 lists soil respiration fluxes reported in other studies of semi-arid environments. Figure 3. Presents representative relations between moisture contents with matric potentials. Figure 4 presents depth profiles on different dates at site TT03 for hydraulic head, matric head, pore space volumetric air fraction, temperature, and the effective CO2 diffusion coefficient. Figure 5a presents the time-dependence of relative water table elevation (relative to the deepest measured depth during the study), daily CO2 flux rates at the soil surface, and soil temperature at a depth of 0.20 m. Figure 5b presents continuous measurements of CO2 flux and soil temperature. Figure 6 presents depth profiles on different dates of CO2 concentrations in the gas phase at the 5 monitoring stations. Figure 7 presents CO2 surface fluxes calculated by the gradient method, comparisons between calculated and measured rates at the soil surface, CO2 fluxes at 2m depth, and the ratio of deep (2 m) to surface CO2 fluxes.

54 ENVIRONMENTAL SCIENCES↗

Grid-Interactive Multi-Zone Building Control Using Reinforcement Learning with Global-Local Policy Search

In this paper, we develop a grid-interactive multi-zone building controller based on a deep reinforcement learning (RL) approach. The controller is designed to facilitate building operation during normal conditions and demand response events, while ensuring occupants comfort and energy efficiency. We leverage a continuous action space RL formulation, and devise a two-stage global-local RL training framework. In the first stage, a global fast policy search is performed using a gradient-free RL algorithm. In the second stage, a local fine-tuning is conducted using a policy gradient method. In contrast to the state-of-the-art model predictive control (MPC) approach, the proposed RL controller does not require complex computation during real-time operation and can adapt to nonlinear building models. We illustrate the controller performance numerically using a five-zone commercial building.

30 DIRECT ENERGY CONVERSION↗

Double Q-Learning for Radiation Source Detection

Anomalous radiation source detection in urban environments is challenging due to the complex nature of background radiation. When a suspicious area is determined, a radiation survey is usually carried out to search for anomalous radiation sources. To locate the source with high accuracy and in a short time, different survey approaches have been studied such as scanning the area with fixed survey paths and data-driven approaches that update the survey path on the fly with newly acquired measurements. In this work, we propose reinforcement learning as a data-driven approach to conduct radiation detection tasks with no human intervention. A simulated radiation environment is constructed, and a convolutional neural network-based double Q-learning algorithm is built and tested for radiation source detection tasks. Simulation results show that the double Q-learning algorithm can reliably navigate the detector and reduce the searching time by at least 44% compared with traditional uniform search methods and gradient search methods.

36 MATERIALS SCIENCE↗

Diffusion Synthetic Acceleration for High-Order Discontinuous Finite Element SN Transport Schemes and Application to Locally Refined Unstructured Meshes

Diffusion synthetic acceleration (DSA) schemes compatible with adaptive mesh refinement (AMR) grids are derived for the SN transport equations discretized using high-order discontinuous finite elements. These schemes are directly obtained from the discretized transport equations by assuming a linear dependence in angle of the angular flux along with an exact Fick's law and, therefore, are categorized as partially consistent. These schemes are akin to the symmetric interior penalty technique applied to elliptic problems and are all based on a second-order discontinuous finite element discretization of a diffusion equation (as opposed to a mixed or P1 formulation). Therefore, they only have the scalar flux as unknowns. A Fourier analysis has been carried out to determine the convergence properties of the three proposed DSA schemes for various cell optical thicknesses and aspect ratios. Out of the three DSA schemes derived, the modified interior penalty (MIP) scheme is stable and effective for realistic problems, even with distorted elements, but loses effectiveness for some highly heterogeneous configurations. The MIP scheme is also symmetric positive definite and can be solved efficiently with a preconditioned conjugate gradient method. Its implementation in an AMR SN transport code has been performed for both source iteration and GMRes-based transport solves, with polynomial orders up to 4. Numerical results are provided and show good agreement with the Fourier analysis results. Results on AMR grids demonstrate that the cost of DSA can be kept low on locally refined meshes.

42 ENGINEERING↗

Optipack

Used to solve basic unconstrained optimization problems that can produce smooth accurate gradients. These are standard published unconstrained optimization such as are described in the book Nocedal and Wright Numerical Optimization: 2nd edition, 2006. This software is generic to any mathematical optimization problem where smooth derivatives exist. There is no specific connection or mention whatsoever to any specific application, period. You cannot find more general mathematical software.OptiPack contains some simple optimization algorithms built on Thyra. Currently, only a few different nonlinear conjugate gradient methods are implemented.

Bartlett, Roscoe↗

Grid-Interactive Multi-Zone Building Control Using Reinforcement Learning with Global-Local Policy Search: Preprint

In this paper, we develop a grid-interactive multi-zone building controller based on a deep reinforcement learning (RL) approach. The controller is designed to facilitate building operation during normal conditions and demand response events, while ensuring occupants comfort and energy efficiency. We leverage a continuous action space RL formulation, and devise a two-stage global-local RL training framework. In the first stage, a global fast policy search is performed using a gradient-free RL algorithm. In the second stage, a local fine-tuning is conducted using a policy gradient method. In contrast to the state-of-the-art model predictive control (MPC) approach, the proposed RL controller does not require complex computation during real-time operation and can adapt to non-linear building models. We illustrate the controller performance numerically using a five-zone commercial building.

30 DIRECT ENERGY CONVERSION↗

Integrated Design of Downwind Land-Based Wind Turbines using Analytic Gradients

Wind turbines are complex systems where component-level changes can have significant system-level effects. Effective wind turbine optimization generally requires an integrated analysis approach with a large number of design variables. Optimizing across large variable sets is orders of magnitude more efficient with gradient-based methods as compared with gradient-free method, particularly when using exact gradients. We have developed a wind turbine analysis set of over 100 components where 90% of the models provide numerically exact gradients through symbolic differentiation, automatic differentiation, and adjoint methods. This framework is applied to a specific design study focused on downwind land-based wind turbines. Downwind machines are of potential interest for large wind turbines where the blades are often constrained by the stiffness required to prevent a tower strike. The mass of these rotor blades may be reduced by utilizing a downwind configuration where the constraints on tower strike are less restrictive. The large turbines of this study range in power rating from 5-7MW and in diameter from 105m to 175m. The changes in blade mass and power production have important effects on the rest of the system, and thus the nacelle and tower systems are also optimized. For high-speed wind sites, downwind configurations do not appear advantageous. The decrease in blade mass (10%) is offset by increases in tower mass caused by the bending moment from the rotor-nacelle-assembly. For low-wind speed sites, the decrease in blade mass is more significant (25-30%) and shows potential for modest decreases in overall cost of energy (around 1-2%).

17 WIND ENERGY↗

Sensitivity of magnetic islands in permanent magnet stellarators using the gradient and Hessian methods

Stellarator plasmas are known to be very sensitive to perturbations in the magnetic field. The permanent magnet stellarator was in part developed as a solution to high machining tolerances placed on the shape properties of electromagnetic coils in traditional stellarators. However, as a consequence of this high sensitivity to the field structure, sensitivities of permanent magnet stellarator plasmas to perturbations of permanent magnet properties must necessarily be well-understood. The gradient and Hessian matrix methods have been previously demonstrated to be useful sensitivity analysis methods for modular coils. We apply these two methods to the study of island width sensitivities in both the MUSE and PM4STELL permanent magnet stellarator projects. These sensitivity methods were used to determine the relative impacts of permanent magnet parameter perturbations on island widths in the vacuum field approximation of both stellarator equilibria. The square of resonant magnetic field perturbation is used here as a proxy for island width. In particular, gradients of magnetizations of individual magnets were examined in MUSE, as well as gradients of magnet group displacements informed by device design. Three different forms of permanent magnet magnetization perturbations are investigated for MUSE, and the flux surface response to perturbations is demonstrated. The Hessian matrix method is applied to PM4STELL, illustrating the sensitivity of dominant island widths to displacements of toroidal wedge structures. These methods allow for selective direction of experimental resources toward regions of heightened sensitivity, while constraints on less impactful permanent magnet parameters can be relaxed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Separation of colloidal two dimensional materials by density gradient ultracentrifugation

Two-dimensional (2D) materials have been made through various approaches but obtaining monodispersed simply by synthesis optimization gained little success, which highlighted the need for introducing nanoseparation methods. Density gradient ultracentrifugation method has emerged as a versatile and scalable method for sorting colloidal 2D nanomaterials. Isopycnic separation was applied on thickness-dependent separation of graphene nanosheets. And rate-zonal separation, as a more versatile separation method, demonstrated its capability in sorting nanosheets of chemically modified single layered graphene, layered double hydroxide, and even metallic Ag. Establishing such density gradient ultracentrifugation method not only achieves monodispersed nanosheets and provides new opportunities for investigation on size dependent properties of 2D materials, but also makes the surface modification possible by introducing “reaction zones” during sedimentation of the colloids. - Graphical abstract: Two-dimensional (2D) materials have been made through various approaches but obtaining monodispersed simply by synthesis optimization gained little success, which highlighted the need for introducing nanoseparation methods. Density gradient ultracentrifugation method has emerged as a versatile and scalable method for sorting colloidal 2D nanomaterials according to their size of thickness difference. Establishing such density gradient ultracentrifugation method not only achieves monodispersed nanosheets and provides new opportunities for investigation on size dependent properties of 2D materials, but also makes the surface modification possible by introducing “reaction zones” during sedimentation of the colloids. - Highlights: • Density gradient ultracentrifugation was applied on size separation of 2D material. • Isopycnic separation was applied on separation of low density materials. • Rate-zonal separation was applied on separation of large density materials. • Size separation lead to investigation on size dependent properties of 2D material. • Mechanism was investigated by introducing “reaction zones” in the gradient.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Learning viscoelasticity models from indirect data using deep neural networks

In this study, we propose a novel approach to model viscoelasticity materials, where rate-dependent and non-linear constitutive relationships are approximated with deep neural networks. We assume that inputs and outputs of the neural networks are not directly observable, and therefore common training techniques with input–output pairs for the neural networks are inapplicable. To that end, we develop a novel computational approach to both calibrate parametric and learn neural-network-based constitutive relations of viscoelasticity materials from indirect displacement data in the context of multiple-physics systems. We show that limited displacement data holds sufficient information to quantify the viscoelasticity behavior. We formulate the inverse computation – modeling viscoelasticity properties from observed displacement data – as a PDE-constrained optimization problem and minimize the error functional using a gradient-based optimization method. The gradients are computed by a combination of automatic differentiation and implicit function differentiation rules. The effectiveness of our method is demonstrated through numerous benchmark problems in geomechanics and porous media transport.

97 MATHEMATICS AND COMPUTING↗

Novel multimodality segmentation using level sets and Jensen-Rényi divergence

Purpose: Positron emission tomography (PET) is playing an increasing role in radiotherapy treatment planning. However, despite progress, robust algorithms for PET and multimodal image segmentation are still lacking, especially if the algorithm were extended to image-guided and adaptive radiotherapy (IGART). This work presents a novel multimodality segmentation algorithm using the Jensen-Rényi divergence (JRD) to evolve the geometric level set contour. The algorithm offers improved noise tolerance which is particularly applicable to segmentation of regions found in PET and cone-beam computed tomography. Methods: A steepest gradient ascent optimization method is used in conjunction with the JRD and a level set active contour to iteratively evolve a contour to partition an image based on statistical divergence of the intensity histograms. The algorithm is evaluated using PET scans of pharyngolaryngeal squamous cell carcinoma with the corresponding histological reference. The multimodality extension of the algorithm is evaluated using 22 PET/CT scans of patients with lung carcinoma and a physical phantom scanned under varying image quality conditions. Results: The average concordance index (CI) of the JRD segmentation of the PET images was 0.56 with an average classification error of 65%. The segmentation of the lung carcinoma images had a maximum diameter relative error of 63%, 19.5%, and 14.8% when using CT, PET, and combined PET/CT images, respectively. The estimated maximal diameters of the gross tumor volume (GTV) showed a high correlation with the macroscopically determined maximal diameters, with aR{sup 2} value of 0.85 and 0.88 using the PET and PET/CT images, respectively. Results from the physical phantom show that the JRD is more robust to image noise compared to mutual information and region growing. Conclusions: The JRD has shown improved noise tolerance compared to mutual information for the purpose of PET image segmentation. Presented is a flexible framework for multimodal image segmentation that can incorporate a large number of inputs efficiently for IGART.

62 RADIOLOGY AND NUCLEAR MEDICINE↗