Search NASA⌕ Search

SEARCH · Search NASA

Results for “Point 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 1,063 records · Page 59

Impact and mitigation of polarized extragalactic foregrounds on Bayesian cosmic microwave background lensing

Future low-noise cosmic microwave background (CMB) lensing measurements from e.g., CMB-S4 will be polarization dominated, rather than temperature dominated. In this new regime, statistically optimal lensing reconstructions outperform the standard quadratic estimator, but their sensitivity to extragalactic polarized foregrounds has not been quantified. Using realistic simulations of polarized radio and infrared point sources, we show for the first time that optimal Bayesian lensing from a CMB-S4-like experiment is insensitive to the expected level of polarized extragalactic foregrounds after masking, as long as an accurate foreground power spectrum is included in the analysis. For more futuristic experiments where these foregrounds could cause a detectable bias, we propose a new method to jointly fit for lensing and the Poisson foregrounds, generalizing the bias hardening from the standard quadratic estimator to Bayesian lensing.

cosmic microwave background↗

ZEUS: An Efficient GPU Optimization Method Integrating PSO, BFGS, and Automatic Differentiation

We introduce a novel, efficient computational method, ZEUS, for numerical optimization, and provide an open-source implementation. It has four key ingredients: (1) particle swarm optimization (PSO), (2) the use of the Broyden-Fletcher-Goldfarb-Shanno (BFGS) method, (3) automatic differentiation (AD), and (4) GPUs. Our approach addresses the computational challenges inherent in high-dimensional, non-convex optimization problems. In the first phase of the algorithm, we get a potentially good set of starting points using PSO. Thereafter, we run BFGS independently in parallel from these starting points. BFGS is one of the best-performing algorithms for numerical optimization. However, it requires the gradient of the function being optimized. ZEUS integrates automatic differentiation into BFGS thus avoiding the need for the user to calculate derivatives explicitly. The use of GPUs allows ZEUS to speed up the calculations substantially. We carry out systematic studies to explore the trade-offs between the number of PSO iterations taken, starting points, and BFGS iteration depth. We show that a handful of iterations of PSO can improve global convergence when combined with BFGS. We also present performance studies using common test functions. The source code can be found at https://github.com/fnal-numerics/global-optimizer-gpu.

Soos, Dominik [Old Dominion U.]↗

Pressure distribution based optimization of phase-coded acoustical vortices

Based on the acoustic radiation of point source, the physical mechanism of phase-coded acoustical vortices is investigated with formulae derivations of acoustic pressure and vibration velocity. Various factors that affect the optimization of acoustical vortices are analyzed. Numerical simulations of the axial, radial, and circular pressure distributions are performed with different source numbers, frequencies, and axial distances. The results prove that the acoustic pressure of acoustical vortices is linearly proportional to the source number, and lower fluctuations of circular pressure distributions can be produced for more sources. With the increase of source frequency, the acoustic pressure of acoustical vortices increases accordingly with decreased vortex radius. Meanwhile, increased vortex radius with reduced acoustic pressure is also achieved for longer axial distance. With the 6-source experimental system, circular and radial pressure distributions at various frequencies and axial distances have been measured, which have good agreements with the results of numerical simulations. The favorable results of acoustic pressure distributions provide theoretical basis for further studies of acoustical vortices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multidisciplinary Optimization of a Transonic Truss Braced Wing Aircraft with Hybrid-Electric Propulsion

Next generation aircraft concepts have subsystems that are increasingly inter-connected. This necessitates advanced design tools using gradient based optimization to properly account for strong subsystem coupling in these aircraft. Additionally, these design tools require accurate modeling capability to support the increased interest in electrified propulsion. Design studies for these aircraft must consider the trade-offs between electric propulsion and turbojet engines at all points in the flight envelope to determine the optimal balance between propulsion options. This paper describes the development of an electric propulsion subsystem model designed to work within Aviary - an open source aircraft design tool. We will demonstrate the electric propulsion model operating with Aviary by showing results from an optimized Transonic Truss-Braced Wing (TTBW) concept using assisted electric propulsion during climb. First, we present results with and without electrification to show the overall system impacts of hybrid electric propulsion during climb. We also present two distinct battery models compatible with this optimization framework, and compare results using both. Next, we ran the same problem using a slightly different cell type, and demonstrate a considerable change in the result. Finally, we vary the cell energy density of the batteries, and provide an illustrative trend for the system-level impact for improving cell technology.

optimization↗

Identifying native point defect configurations in α-alumina

Intimately intertwined atomic and electronic structures of point defects govern diffusion-limited corrosion and underpin the operation of optoelectronic devices. For some materials, complex energy landscapes containing metastable defect configurations challenge first-principles modeling efforts. Here, we thoroughly reevaluate native point defect geometries for the illustrative case of α-Al 2 O 3 by comparing three methods for sampling candidate geometries in density functional theory calculations: displacing atoms near a naively placed defect, initializing interstitials at high-symmetry points of a Voronoi decomposition, and Bayesian optimization. We find symmetry-breaking distortions for oxygen vacancies in some charge states, and we identify several distinct oxygen split-interstitial geometries that help explain literature discrepancies involving this defect. We also report a surprising and, to our knowledge, previously unknown trigonal geometry favored by aluminum interstitials in some charge states. Importantly, these new configurations may have transformative impacts on our understanding of defect migration pathways in aluminum-oxide scales protecting metal alloys from corrosion. Overall, the Voronoi scheme appears most effective for sampling candidate interstitial sites because it always succeeded in finding the lowest-energy geometry identified in this study, although no approach found every metastable configuration. Finally, we show that the position of defect levels within the band gap can depend strongly on the defect geometry, underscoring the need to conduct careful searches for ground-state geometries in defect calculations.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Development of a Vibration Technique Based on Geometric Optimization for Fatigue Life Evaluation of Sandwich Composite Structures

A major obstacle to obtaining cost-effective experimental data on the fatigue life of sandwich panels is the prohibitive amount of time and cost required to carry out millions of cycles. On the other hand, vibration techniques applied to sandwich geometries fail to match the stress patterns that are obtained from standard flexural fatigue tests. To overcome such limitations, a vibration-based fatigue technique is proposed, which entails the use of sandwich specimens whose geometries are optimized to reproduce the stress distribution observed during three point bend loading while vibrating at the first resonant frequency. The proposed vibration technique was experimentally validated. The results, compared with the average number of cycles to failure at different stress ratios obtained via the Three-Point Bending test, showed high levels of accuracy. The proposed method is robust and time effective and indicates the possibility of attaining fatigue lifetime prediction of a wide class of composite elements, such as sandwich panels.

36 MATERIALS SCIENCE↗

Gas-kinetic unified algorithm for hypersonic flows covering various flow regimes solving Boltzmann model equation in nonequilibrium effect

Based on the Gas-Kinetic Unified Algorithm (GKUA) directly solving the Boltzmann model equation, the effect of rotational non-equilibrium is investigated recurring to the kinetic Rykov model with relaxation property of rotational degrees of freedom. The spin movement of diatomic molecule is described by moment of inertia, and the conservation of total angle momentum is taken as a new Boltzmann collision invariant. The molecular velocity distribution function is integrated by the weight factor on the internal energy, and the closed system of two kinetic controlling equations is obtained with inelastic and elastic collisions. The optimization selection technique of discrete velocity ordinate points and numerical quadrature rules for macroscopic flow variables with dynamic updating evolvement are developed to simulate hypersonic flows, and the gas-kinetic numerical scheme is constructed to capture the time evolution of the discretized velocity distribution functions. The gas-kinetic boundary conditions in thermodynamic non-equilibrium and numerical procedures are studied and implemented by directly acting on the velocity distribution function, and then the unified algorithm of Boltzmann model equation involving non-equilibrium effect is presented for the whole range of flow regimes. The hypersonic flows involving non-equilibrium effect are numerically simulated including the inner flows of shock wave structures in nitrogen with different Mach numbers of 1.5-Ma-25, the planar ramp flow with the whole range of Knudsen numbers of 0.0009-Kn-10 and the three-dimensional re-entering flows around tine double-cone body.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Deep reinforcement learning assisted co-optimization of Volt-VAR grid service in distribution networks

With the increasing penetration of distributed energy resources in distribution networks, Volt-VAR control and optimization (VVC/VVO) have become very important to ensure an acceptable quality of service to all customers. System operators can rely on slow-responding utility devices, including capacitor banks and on-load tap changing transformers, along with fast-responding battery and photovoltaic (PV) inverters for the VVC/VVO implementation. Because of variations in response time of these two classes of devices, and different control actions (discrete versus continuous), coordinated and optimal scheduling and operation have become of utmost importance. Here, this paper develops a look-ahead deep reinforcement learning (DRL)-based multi-objective VVO technique to improve the voltage profile of active distribution networks, decrease network and inverter power loss, and save the operational cost of the grid. It proposes a deep deterministic policy gradient (DDPG)-based approach to schedule the optimal reactive and/or active power set-points of fast-responding inverters, and a deep Q-network (DQN)-based DRL agent to schedule the discrete decisions variables of slow-responding assets. The reactive power output of PV and battery smart inverters are scheduled at 30-minute intervals and the capacitors’ commitment status is scheduled with several hour intervals. The proposed framework is validated on the modified IEEE 34-bus and 123-bus test cases with embedded PV and PV-plus-storage. To validate the efficacy of the proposed VVO, it is compared with several scenarios, including the base case without VVO, localized droop control of DERs, DDPG-only, and twin delayed DDPG (TD3) agent-based DRL techniques. The results justify the superior performance of the proposed method to improve the voltage profile, reduce network power loss, and minimize the look-ahead grid operational cost while minimizing the undesirable power losses in inverters as a result of power factor adjustments.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Tracking Excited States in Wave Function Optimization Using Density Matrices and Variational Principles

We introduce an approach for finding individual excited states' energy stationary points in complete active space self-consistent field theory that is compatible with standard optimization methods and highly effective at overcoming difficulties due to root flipping and near-degeneracies. Inspired by both the maximum overlap method and recent progress in excited-state variational principles, our approach combines these ideas in order to track individual excited states throughout the orbital optimization process. In a series of tests involving root flipping, near-degeneracies, charge transfers, and double excitations, we show that this approach is more effective for state-specific optimization than either the naive selection of roots on the basis of energy ordering or a more direct generalization of the maximum overlap method. We offer evidence that this state-specific approach improves the performance of complete active space perturbation theory for vertical excitation energies. Furthermore, we find that the state-specific optimization can help avoid state-averaging-induced discontinuities on potential energy surfaces. With a simple implementation, a low cost, and compatibility with large active space methods, the approach is designed to be useful in a wide range of excited-state investigations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Coordination of Behind-the-Meter Energy Storage and Building Loads: Optimization with Deep Learning Models*

With the increasing penetration of renewable energy systems and energy storage systems in buildings, it is critical to optimize system operation. In this paper, we develop an optimization strategy to minimize the operation cost as well as maintain indoor thermal comfort, for a building integrated with battery and PV. A Recurrent Neural Network (RNN) is used to predict building thermal load and zone temperatures. A black-box optimization algorithm known as nonlinear optimization by mesh adaptive direct search (NOMADS) is employed in simulation to provide look-ahead optimal battery dispatch and zone temperature set-point schedules so that the operation cost is minimized. Field data collected from a medium sized office building at Pacific Northwest National Laboratory (PNNL) integrated with Photovoltaic (PV) system model and battery storage system model is used to demonstrate the proposed methodology. Compared with rule based methods, the optimization strategy obtained a lower cost of operation while satisfying comfort constraints.

Chen, Yimin↗

Short-Term Load Forecasting-Based Automatic Distribution Network Reconfiguration

In a traditional dynamic network reconfiguration study, the optimal topology is determined at every scheduled time point by using the real load data measured at that time. The development of the load forecasting technique can provide an accurate prediction of the load power that will happen in a future time and provide more information about load changes. With the inclusion of load forecasting, the optimal topology can be determined based on the predicted load conditions during a longer time period instead of using a snapshot of the load at the time when the reconfiguration happens; thus, the distribution system operator can use this information to better operate the system reconfiguration and achieve optimal solutions. This paper proposes a short-term load forecasting approach to automatically reconfigure distribution systems in a dynamic and pre-event manner. Specifically, a short-term and high-resolution distribution system load forecasting approach is proposed with a forecaster based on support vector regression and parallel parameters optimization. The network reconfiguration problem is solved by using the forecasted load continuously to determine the optimal network topology with the minimum amount of loss at the future time. The simulation results validate and evaluate the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Bond-associated deformation gradients for peridynamic correspondence model

Non-ordinary state-based peridynamic correspondence material model is known to have issues with material instability, i.e. the existence of zero-energy modes, due to non-unique mapping between deformation states and force states via the conventional peridynamic deformation gradient. In this paper, an alternative approach in which the deformation gradient hence force state are computed specifically for each individual bond is proposed to eliminate the material instability. Bond-associated deformation gradient is calculated based on deformation states of material points within an individual bond’s proximity, termed here as the bond-associated family, rather than a material point’s whole family. This bond-associated deformation gradient can better represent the force state of each individual bond from the deformation states of material points in its proximity, and hence inherently resolves issues of material instability in the conventional correspondence material model. Parametric study on bond-associated horizon size indicates that the optimal size should be no less than the material point’s horizon size but smaller than two times of that value. Comparisons against reference solutions using finite element method establish the validity and accuracy of the proposed formulation.

42 ENGINEERING↗

Operational results from the LHC luminosity monitors

The luminosity monitors for the high luminosity regions in the LHC have been operating to monitor and optimize the luminosity since 2009. The device is a gas ionization chamber inside the neutral particle absorber 140 m from the interaction point and monitors showers produced by high energy neutral particles from the collisions. It has the ability to resolve the bunch-by-bunch luminosity as well as to survive the extreme level of radiation in the nominal LHC operation. We present operational results of the device during proton and lead ion operations in 2010 and make comparisons with measurements of experiments. The Large Hadron Collider (LHC) at CERN can accelerate proton and lead ion beams to 7 TeV and 547 TeV and produce collisions of these particles. Luminosity measures performance of the LHC and is particularly important for experiments in high luminosity interaction points (IPs), ATLAS (IP1) and CMS (IP5). To monitor and optimize the luminosities of these IPs, BRAN (Beam RAte Neutral) detectors [1, 2] have been installed and operating since the beginning of the 2009 operation [3]. A neutral particle absorber (TAN) protects the D2 separation dipole from high energy forward neutral particles produced in the collisions [4]. These neutral particles produce electromagnetic and hadronic showers inside the TAN and their energy flux is proportional to the collision rate and hence to the luminosity. The BRAN detector is an Argon gas ionization chamber installed inside the TANs on both sides of the IP1 and IP5 and monitors the relative changes in the luminosity by detecting the ionization due to these showers. When the number of collisions per bunch crossing (multiplicity) is small, the shower rate inside the TAN is also proportional to the luminosity. Hence, the detector is designed to operate by measuring either the shower rate (counting mode for low and intermediate luminosities) or the average shower flux (pulse height mode for high luminosities). The detector is also designed (1) to survive the extreme level of radiation ({approx}1 GGy in the nominal condition), (2) to resolve the shower from each bunch crossing (40 MHz in the nominal condition) and measure the bunch-by-bunch luminosities, and (3) to have four independent square shaped channels, each occupying a quadrant, making the detector sensitive to the crossing angle [1, 2]. During the proton operation in 2010, the beam energy was 3.5 TeV and the multiplicity did not exceed four. Because the counting mode is still effective in such a condition [5], the BRAN were operated in the counting mode in 2010. This paper presents operational results of the BRANs during the operation in 2010 (mainly the proton operation) and makes comparisons with measurements of the experiments. The luminosity optimization is discussed in detail in [6] and so this paper focuses on measurements during the normal operation.

43 PARTICLE ACCELERATORS↗

Effect of Electrode Geometry on the Classification Performance of Rapid Evaporative Ionization Mass Spectrometric (REIMS) Bacterial Identification

The recently developed automated, high-throughput monopolar REIMS platform is suited for the identification of clinically important microorganisms. Although already comparable to the previously reported bipolar forceps method, optimization of the geometry of monopolar electrodes, at the heart of the system, holds the most scope for further improvements to be made. For this, sharp tip and round shaped electrodes were optimized to maximize species-level classification accuracy. Following optimization of the distance between the sample contact point and tube inlet with the sharp tip electrodes, the overall cross-validation accuracy improved from 77% to 93% in negative and from 33% to 63% in positive ion detection modes, compared with the original 4 mm distance electrode. As an alternative geometry, round tube shaped electrodes were developed. Geometry optimization of these included hole size, number, and position, which were also required to prevent plate pick-up due to vacuum formation. Additional features, namely a metal “X”-shaped insert and a pin in the middle were included to increase the contact surface with a microbial biomass to maximize aerosol production. Following optimization, cross-validation scores showed improvement in classification accuracy from 77% to 93% in negative and from 33% to 91% in positive ion detection modes. Supervised models were also built, and after the leave 20% out cross-validation, the overall classification accuracy was 98.5% in negative and 99% in positive ion detection modes. This suggests that the new generation of monopolar REIMS electrodes could provide substantially improved species level identification accuracies in both polarity detection modes. .

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Convex Relaxations in Power System Optimization, A Brief Introduction

Convex relaxations of the AC power ow equations have attracted signi cant interest in the power systems research community in recent years. The following collection of video lectures provides a brief introduction to the mathematics of AC power systems, continuous nonlinear optimization, and relaxations of the power ow equations. The aim of the videos is to provide the high level ideas of convex relaxations and their applications in power system optimization, and could be used as a starting point for researchers who want to study, use or develop new convex relaxations for use in their own research. The videos do not aim to provide an in-depth tutorial about speci c convex relaxations, but rather focus on ideas that are common to all convex relaxations of the AC optimal power ow problem. The videos assume minimal prerequisite knowledge and are designed to be accessible to a variety of disciplines (e.g. power system engineers, industrial engineers and computer scientists). They are designed to be modular, such that interested listeners can either watch the series as a whole, or choose only the subset that matches their interest. The following sections provide a brief overview of the topics and references covered in each video.

97 MATHEMATICS AND COMPUTING↗

Applications Performance Under MPL and MPI on NAS IBM SP2

On July 5, 1994, an IBM Scalable POWER parallel System (IBM SP2) with 64 nodes, was installed at the Numerical Aerodynamic Simulation (NAS) Facility Each node of NAS IBM SP2 is a "wide node" consisting of a RISC 6000/590 workstation module with a clock of 66.5 MHz which can perform four floating point operations per clock with a peak performance of 266 Mflop/s. By the end of 1994, 64 nodes of IBM SP2 will be upgraded to 160 nodes with a peak performance of 42.5 Gflop/s. An overview of the IBM SP2 hardware is presented. The basic understanding of architectural details of RS 6000/590 will help application scientists the porting, optimizing, and tuning of codes from other machines such as the CRAY C90 and the Paragon to the NAS SP2. Optimization techniques such as quad-word loading, effective utilization of two floating point units, and data cache optimization of RS 6000/590 is illustrated, with examples giving performance gains at each optimization step. The conversion of codes using Intel's message passing library NX to codes using native Message Passing Library (MPL) and the Message Passing Interface (NMI) library available on the IBM SP2 is illustrated. In particular, we will present the performance of Fast Fourier Transform (FFT) kernel from NAS Parallel Benchmarks (NPB) under MPL and MPI. We have also optimized some of Fortran BLAS 2 and BLAS 3 routines, e.g., the optimized Fortran DAXPY runs at 175 Mflop/s and optimized Fortran DGEMM runs at 230 Mflop/s per node. The performance of the NPB (Class B) on the IBM SP2 is compared with the CRAY C90, Intel Paragon, TMC CM-5E, and the CRAY T3D.

Saini, Subhash↗

The Newton Modified Barrier Method for QP Problems

The Modified Barrier Functions (MBF) have elements of both Classical Lagrangians (CL) and Classical Barrier Functions (CBF). The MBF methods find an unconstrained minimizer of some smooth barrier function in primal space and then update the Lagrange multipliers, while the barrier parameter either remains fixed or can be updated at each step. The numerical realization of the MBF method leads to the Newton MBF method, where the primal minimizer is found by using Newton's method. This minimizer is then used to update the Lagrange multipliers. In this paper, we examine the Newton MBF method for the Quadratic Programming (QP) problem. It will be shown that under standard second-order optimality conditions, there is a ball around the primal solution and a cut cone in the dual space such that for a set of Lagrange multipliers in this cut cone, the method converges quadratically to the primal minimizer from any point in the aforementioned ball, and continues to do so after each Lagrange multiplier update. The Lagrange multipliers remain within the cut cone and converge linearly to their optimal values. Any point in this ball will be called a "hot start". Starting at such a "hot start", at most Omicron(1n 1n epsilon(exp -1)) Newton steps are sufficient to perform the primal minimization which is necessary for the Lagrange multiplier update. Here, epsilon > 0 is the desired accuracy. Because of the linear convergence of the Lagrange multipliers, this means that only Omicron(1n epsilon(exp -1))omicron(ln 1n epsilon(exp-1)) Newton steps are required to reach an epsilon-approximation to the solution from any "hot start". In order to reach the "hot start", one has to perform Omicron(square root(m) 1n C) Newton steps, where m characterizes the size of the problem and C > 0 is the condition number of the QP problem. This condition number will be characterized explicitly in terms of key parameters of the QP problem, which in turn depend on the input data and the size of the problem.

Melman, A.↗