Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ANDERSON”

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 163 records · Page 9

Anderson acceleration stability in NDA-accelerated k-eigenvalue problems

Anderson acceleration (AA) has been used to improve the stability and convergence rate of multiphysics iterative methods for reactor analysis. Most applications studied assume a tightly converged solution for the different physics problems, and AA is usually applied to state variables like temperature, density, and heat generation rate. In this paper, we study the theoretical performance of AA in NDA-accelerated k-eigenvalue problems. The problems and algorithms studied are simplified from the coupled iteration scheme adopted by MPACT and many other high-fidelity whole-core reactor codes. Compared to previous analyses of AA for these iteration schemes, we study the case with a partially converged neutronics solution and possibly partially converged nonlinear diffusion acceleration (NDA)/coarse mesh finite difference (CMFD) solutions. We observe that the performance of the iteration scheme with AA is very sensitive to the initial guess and is affected by the partially converged CMFD solutions. When the NDA solution is fully converged, using AA cannot achieve the optimal convergence rate in large-sized problems. Conversely, if the NDA solution is partially converged, the iteration scheme with AA can diverge or converge extremely slowly. It is found that the loss of robustness for AA is due to the fact that it is applied to the iterative subspace of state variables rather than the fundamental unknowns of the governing equations. To improve the robustness, the scalar flux should also be considered in the implementation of AA. After considering the residuals of flux, we observe that the stability is regardless of the partial convergence of NDA solutions. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

FedOSAA: Improving Federated Learning with One-Step Anderson Acceleration

Federated learning (FL) is a distributed machine learning approach that enables multiple local clients and a central server to collaboratively train a model while keeping the data on their own devices. First-order methods, particularly those incorporating variance reduction techniques, are the most widely used FL algorithms due to their simple implementation and stable performance. However, these methods tend to be slow and require a large number of communication rounds to reach the global minimizer. We propose FedOSAA, a novel approach that preserves the simplicity of first-order methods while achieving the rapid convergence typically associated with second-order methods. Our approach applies one Anderson acceleration (AA) step following classical local updates based on first-order methods with variance reduction, such as FedSVRG and SCAFFOLD, during local training. This AA step is able to leverage curvature information from the history points and gives a new update that approximates the Newton-GMRES direction, thereby significantly improving the convergence. We establish a local linear convergence rate to the global minimizer of FedOSAA for smooth and strongly convex loss functions. Numerical comparisons show that FedOSAA substantially improves the communication and computation efficiency of the original first-order methods, achieving performance comparable to second-order methods like GIANT.

Feng, Xue [University of California, Davis]↗

Extended recursion in operator space (EROS), a new impurity solver for the single impurity Anderson model

We have developed a new efficient and accurate impurity solver for the single impurity Anderson model (SIAM), which is based on a non-perturbative recursion technique in a space of operators and involves expanding the self-energy as a continued fraction. The method has no special occupation number or temperature restrictions; the only approximation is the number of levels of the continued fraction retained in the expansion. We also show how this approach can be used as a new approach to Dynamical Mean Field Theory (DMTF) and illustrate this with the Hubbard model. The three lowest orders of recursion give the Hartree-Fock, Hubbard I, and Hubbard III approximations. A higher level of recursion is able to reproduce the expected 3-peak structure in the spectral function and Fermi liquid behavior.

66↗

GW approach to Anderson model in and out of equilibrium : scaling properties in the Kondo regime.

The low-energy properties of the Anderson model for a single impurity coupled to two leads are studied using the GW approximation. We find that quantities such as the spectral function at zero temperature, the linear-response conductance as function of temperature or the differential conductance as function of bias voltage exhibit universal scaling behavior in the Kondo regime. We show how the form of the GW scaling functions relates to the form of the scaling functions obtained from the exact solution at equilibrium. We also compare the energy scale that goes inside the GW scaling functions with the exact Kondo temperature, for a broad range of the Coulomb interaction strength in the asymptotic regime. This analysis allows to clarify a presently suspended question in the literature, namely whether or not the GW solution captures the Kondo resonance.

97 MATHEMATICS AND COMPUTING↗

Weak chaos in the disordered nonlinear Schroedinger chain: Destruction of Anderson localization by Arnold diffusion

Research Highlights: > In a one-dimensional disordered chain of oscillators all normal modes are localized. > Nonlinearity leads to chaotic dynamics. > Chaos is concentrated on rare chaotic spots. > Chaotic spots drive energy exchange between oscillators. > Macroscopic transport coefficients are obtained. - Abstract: The subject of this study is the long-time equilibration dynamics of a strongly disordered one-dimensional chain of coupled weakly anharmonic classical oscillators. It is shown that chaos in this system has a very particular spatial structure: it can be viewed as a dilute gas of chaotic spots. Each chaotic spot corresponds to a stochastic pump which drives the Arnold diffusion of the oscillators surrounding it, thus leading to their relaxation and thermalization. The most important mechanism of equilibration at long distances is provided by random migration of the chaotic spots along the chain, which bears analogy with variable-range hopping of electrons in strongly disordered solids. The corresponding macroscopic transport equations are obtained.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Long-Term Results of the M. D. Anderson Randomized Dose-Escalation Trial for Prostate Cancer

Purpose: To report the long-term results of a randomized radiotherapy dose escalation trial for prostate cancer. Methods and Materials: From 1993 to 1998, a total of 301 patients with stage T1b to T3 prostate cancer were accrued to a randomized external beam dose escalation trial using 70 Gy versus 78 Gy. The median follow-up is now 8.7 years. Kaplan-Meier analysis was used to compute rates of prostate-specific antigen (PSA) failure (nadir + 2), clinical failure, distant metastasis, disease-specific, and overall survival as well as complication rates at 8 years post-treatment. Results: For all patients, freedom from biochemical or clinical failure (FFF) was superior for the 78-Gy arm, 78%, as compared with 59% for the 70-Gy arm (p = 0.004, and an even greater benefit was seen in patients with initial PSA >10 ng/ml (78% vs. 39%, p = 0.001). The clinical failure rate was significantly reduced in the 78-Gy arm as well (7% vs. 15%, p = 0.014). Twice as many patients either died of prostate cancer or are currently alive with cancer in the 70-Gy arm. Gastrointestinal toxicity of grade 2 or greater occurred twice as often in the high dose patients (26% vs. 13%), although genitourinary toxicity of grade 2 or greater was less (13% vs. 8%) and not statistically significantly different. Dose-volume histogram analysis showed that the complication rate could be significantly decreased by reducing the amount of treated rectum. Conclusions: Modest escalation in radiotherapy dose improved freedom from biochemical and clinical progression with the largest benefit in prostate cancer patients with PSA >10 ng/ml.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Long-Term Clinical Outcome of Intensity-Modulated Radiotherapy for Inoperable Non-Small Cell Lung Cancer: The MD Anderson Experience

Purpose: In 2007, we published our initial experience in treating inoperable non-small-cell lung cancer (NSCLC) with intensity-modulated radiation therapy (IMRT). The current report is an update of that experience with long-term follow-up. Methods and Materials: Patients in this retrospective review were 165 patients who began definitive radiotherapy, with or without chemotherapy, for newly diagnosed, pathologically confirmed NSCLC to a dose of {>=}60 Gy from 2005 to 2006. Early and late toxicities assessed included treatment-related pneumonitis (TRP), pulmonary fibrosis, esophagitis, and esophageal stricture, scored mainly according to the Common Terminology Criteria for Adverse Events 3.0. Other variables monitored were radiation-associated dermatitis and changes in body weight and Karnofsky performance status. The Kaplan-Meier method was used to compute survival and freedom from radiation-related acute and late toxicities as a function of time. Results: Most patients (89%) had Stage III to IV disease. The median radiation dose was 66 Gy given in 33 fractions (range, 60-76 Gy, 1.8-2.3 Gy per fraction). Median overall survival time was 1.8 years; the 2-year and 3-year overall survival rates were 46% and 30%. Rates of Grade {>=}3 maximum TRP (TRP{sub max}) were 11% at 6 months and 14% at 12 months. At 18 months, 86% of patients had developed Grade {>=}1 maximum pulmonary fibrosis (pulmonary fibrosis{sub max}) and 7% Grade {>=}2 pulmonary fibrosis{sub max}. The median times to maximum esophagitis (esophagitis{sub max}) were 3 weeks (range, 1-13 weeks) for Grade 2 and 6 weeks (range, 3-13 weeks) for Grade 3. A higher percentage of patients who experienced Grade 3 esophagitis{sub max} later developed Grade 2 to 3 esophageal stricture. Conclusions: In our experience, using IMRT to treat NSCLC leads to low rates of pulmonary and esophageal toxicity, and favorable clinical outcomes in terms of survival.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Anderson critical metal phase in trivial states protected by average magnetic crystalline symmetry

Transitions between distinct obstructed atomic insulators (OAIs) protected by crystalline symmetries, where electrons form molecular orbitals centering away from the atom positions, must go through an intermediate metallic phase. In this work, we find that the intermediate metals will become a scale-invariant critical metal phase (CMP) under certain types of quenched disorder that respect the magnetic crystalline symmetries on average. We explicitly construct models respecting average C 2z T, m, and C 4z T and show their scale-invariance under chemical potential disorder by the finite-size scaling method. Conventional theories, such as weak anti-localization and topological phase transition, cannot explain the underlying mechanism. A quantitative mapping between lattice and network models shows that the CMP can be understood through a semi-classical percolation problem. Ultimately, we systematically classify all the OAI transitions protected by (magnetic) groups Pm, P2', P4', and P6' with and without spin-orbit coupling, most of which can support CMP.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Hanbury Brown and Twiss correlations of Anderson localized waves

When light waves propagate through disordered photonic lattices, they can eventually become localized due to multiple scattering effects. Here we show experimentally that while the evolution and localization of the photon density distribution is similar in the two cases of diagonal and off-diagonal disorder, the density-density correlation carries a distinct signature of the type of disorder. We show that these differences reflect a symmetry in the spectrum and eigenmodes that exists in off-diagonally disordered lattices but is absent in lattices with diagonal disorder.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Proposed signature of Anderson localization and correlation-induced delocalization in an N-leg optical lattice

We propose a realization of the one-dimensional random dimer model and certain N-leg generalizations using cold atoms in an optical lattice. We show that these models exhibit multiple delocalization energies that depend strongly on the symmetry properties of the corresponding Hamiltonian, and we provide analytical and numerical results for the localization length as a function of energy. We demonstrate that the N-leg systems possess similarities with their one-dimensional ancestors but are demonstrably distinct. The existence of critical delocalization energies leads to dips in the momentum distribution that serve as a clear signal of the localization-delocalization transition. These momentum distributions are different for models with different group symmetries and are identical for those with the same symmetry.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Anderson localization in optical lattices with speckle disorder

We study the localization properties of noninteracting waves propagating in a speckle-like potential superposed on a one-dimensional lattice. Using a combined decimation-renormalization procedure, we estimate the localization length for a tight-binding Hamiltonian where site energies are square-sinc-correlated random variables. By decreasing the width of the correlation function, the disorder patterns approach a {delta}-correlated disorder, and the localization length becomes almost energy independent in the strong disorder limit. We show that this regime can be reached for a size of the speckle grains on the order of (lower than) four lattice steps.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗