Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “optimization 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 505 records · Page 28
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

A multi-perspective review of microbial fuel-cells for wastewater treatment: Bio-electro-chemical, microbiologic and modeling aspects

Microbial Fuel Cells (MFCs) represent a still novel technology for the recovery of energy and resources through wastewater treatment. Although the technology is quite appealing, due its potential benefits, its practical application is still hampered by several drawbacks, such as systems instability (especially when attempting to scale-up reactors from laboratory prototype), internally competing microbial reactions, and limited power generation. This paper is an attempt to address several of the operational issues related to MFCs application to wastewater treatment, in particular when dealing with simultaneous organic matter and nitrogen pollution control. Reactor configuration, operational schemes, electrochemical and microbiological characterization, optimization methods and modelling strategies are reviewed and discussed with a multidisciplinary, multi-perspective approach. The conclusions drawn herein can be of practical interest for all MFC researchers dealing with domestic or industrial wastewater treatment..

09 BIOMASS FUELS↗
DOE OSTI2020

Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations

Here we propose a generalized space-time domain decomposition approach for the physics-informed neural networks (PINNs) to solve nonlinear partial differential equations (PDEs) on arbitrary complex-geometry domains. The proposed framework, named eXtended PINNs ( X P I N N s ), further pushes the boundaries of both PINNs as well as conservative PINNs (cPINNs), which is a recently proposed domain decomposition approach in the PINN framework tailored to conservation laws. Compared to PINN, the XPINN method has large representation and parallelization capacity due to the inherent property of deployment of multiple neural networks in the smaller subdomains. Unlike cPINN, XPINN can be extended to any type of PDEs. Moreover, the domain can be decomposed in any arbitrary way (in space and time), which is not possible in cPINN. Thus, XPINN offers both space and time parallelization, thereby reducing the training cost more effectively. In each subdomain, a separate neural network is employed with optimally selected hyperparameters, e.g., depth/width of the network, number and location of residual points, activation function, optimization method, etc. A deep network can be employed in a subdomain with complex solution, whereas a shallow neural network can be used in a subdomain with relatively simple and smooth solutions. We demonstrate the versatility of XPINN by solving both forward and inverse PDE problems, ranging from one-dimensional to three-dimensional problems, from time-dependent to time-independent problems, and from continuous to discontinuous problems, which clearly shows that the XPINN method is promising in many practical problems. The proposed XPINN method is the generalization of PINN and cPINN methods, both in terms of applicability as well as domain decomposition approach, which efficiently lends itself to parallelized computation. The XPINN code is available on h t t p s : / / g i t h u b . c o m / A m e y a J a g t a p / X P I N N s .

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Targeted Metabolomics of the Phenylpropanoid Pathway in Arabidopsis thaliana using Reversed Phase Liquid Chromatography Coupled with Tandem Mass Spectrometry: Measuring Lignin Precursors by LC-MS

The phenylpropanoid pathway is a source of a diverse group of compounds derived from phenylalanine, many of which are involved in lignin biosynthesis and serve as precursors for the production of valuable compounds, such as coumarins, flavonoids, and lignans. Consequently, recent efforts have been invested in mechanistically understanding monolignol biosynthesis, making the quantification of these metabolites vital. To develop an improved and comprehensive analytical method for (i) extensively profiling, and (ii) accurately quantifiying intermediates of the monolignol biosynthetic network, using Arabidopsis thaliana as a model system. A liquid chromatography-tandem mass spectrometry with electrospray ionization was developed to quantify phenylpropanoid metabolites in Arabidopsis wildtype and cinnamoyl CoA reductase1 (CCR1) deficient lines (ccr1). Vortexing at high temperatures (65°C) enhanced release of phenylpropanoids, specifically the more hydrophobic compounds. A pH of 5.3 and ammonium acetate buffer concentration of 2.5 mM resulted in an optimal analyte response across standards. Ion suppression was estimated using standard spike recovery studies for accurate quantitation. The optimized method was used to profile Arabidopsis wildtype and ccr1 stems. An increase in hydroxycinnamic acid derivatives and a decrease in the hydroxycinnamyl aldehydes and alcohols in ccr1 lines, supports a shift of flux from lignin synthesis to other secondary metabolites and phenylpropanoid derivatives. Compared to existing targeted profiling techniques, our method is capable of quantifying a wider range of intermediates (15 out of 22 in WT Arabidopsis stems) at low in vivo concentrations (~50 pmol/g-FW for certain compounds), while requiring minimal sample preparation.

09 BIOMASS FUELS↗
DOE OSTI2015

Optical modeling and electrical properties of cadmium oxide nanofilms: Developing a meta–heuristic calculation process model

Cadmium oxide thin films were deposited onto glass substrates by sol–gel dip-coating method and annealed in air. The normal incidence transmittance of the films was measured by a spectrophotometer. D.C electrical parameters such as carrier concentration and mobility were analyzed by Hall Effect measurements. A combination of Forouhi–Bloomer and standard Drude model was used to simulate the optical constants and thicknesses of the films from transmittance data. The transmittance spectra of the films in the visible domain of wavelengths were successfully fitted by using the result of a hybrid particle swarm optimization method and genetic algorithm. The simulated transmittance is in good accordance with the measured spectrum in the whole measurement wavelength range. The electrical parameters obtained from the optical simulation are well consistent with those measured electrically by Hall Effect measurements.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

Reliability‐based layout optimization in offshore wind energy systems

Abstract Existing methods for optimizing wind array layouts typically use power or cost objectives and rarely consider reliability‐based objectives. Component and system failure rates, however, are dependent on location‐specific wind conditions, are influenced by array layout and wake interactions, and have a direct and significant impact on capital costs, operational costs, and power production. Although wind power plant models exist that calculate wind loads with sufficient resolution to capture component loading dynamics from wind conditions, they are computationally expensive and thus not suitable for research applications requiring many evaluations, particularly optimization. This study describes the development of computationally efficient, reliability‐based layout optimization methods, enabling us to explore the relationship between component reliability and layout optimization. These methods include the surrogate modeling of the planet bearing life based on varying wind conditions simulated in FAST.Farm and the formulation of reliability‐based objectives based on failure cost and power production models. Through demonstration of this method, we explore how wind conditions, objective functions, and capacity density influence reliability‐based layout optimization. Results indicate that considering reliability alongside power production can reduce failure costs associated with replacement costs and downtime while maintaining or improving power production. Our conclusions highlight the opportunity for wind power plant developers to integrate reliability and operational expenditures alongside performance and capital expenditure objectives in plant design and development to improve plant performance and costs.

17 WIND ENERGY↗
DOE OSTI2023

BayFlux: A Bayesian method to quantify metabolic Fluxes and their uncertainty at the genome scale

Metabolic fluxes, the number of metabolites traversing each biochemical reaction in a cell per unit time, are crucial for assessing and understanding cell function. 13 C Metabolic Flux Analysis ( 13 C MFA) is considered to be the gold standard for measuring metabolic fluxes. 13 C MFA typically works by leveraging extracellular exchange fluxes as well as data from 13 C labeling experiments to calculate the flux profile which best fit the data for a small, central carbon, metabolic model. However, the nonlinear nature of the 13 C MFA fitting procedure means that several flux profiles fit the experimental data within the experimental error, and traditional optimization methods offer only a partial or skewed picture, especially in “non-gaussian” situations where multiple very distinct flux regions fit the data equally well. Here, we present a method for flux space sampling through Bayesian inference (BayFlux), that identifies the full distribution of fluxes compatible with experimental data for a comprehensive genome-scale model. This Bayesian approach allows us to accurately quantify uncertainty in calculated fluxes. We also find that, surprisingly, the genome-scale model of metabolism produces narrower flux distributions (reduced uncertainty) than the small core metabolic models traditionally used in 13 C MFA. The different results for some reactions when using genome-scale models vs core metabolic models advise caution in assuming strong inferences from 13 C MFA since the results may depend significantly on the completeness of the model used. Based on BayFlux, we developed and evaluated novel methods (P- 13 C MOMA and P- 13 C ROOM) to predict the biological results of a gene knockout, that improve on the traditional MOMA and ROOM methods by quantifying prediction uncertainty.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2014

TH-A-BRE-01: The Status of Intensity Modulated Proton and Ion Therapy

IMRT with photons has become a radiation therapy standard of care for many cancer treatment sites. The situation is quite different with intensity modulated particle (protons and ion) radiation therapy (IMPT). With the rapid development of beam scanning techniques and many of the newer proton facilities exclusively offering active beam scanning as their radiation delivery technique, it is timely to give an update on the status and challenges of IMPT. The leading principle in IMPT is to aim at the target from several, not necessarily coplanar, directions with multiple pencil beams that are modulated in their intensity and adjusted in their energy such that a desired dose distribution or, more generally, a desired bio-effective dose distribution is achieved. Different from low-LET photons, the varying relative biological effectiveness (RBE) along the beam path adds an additional dimension to the treatment planning process and will require biophysical modeling at least for carbon ion therapy. IMPT involves computationally challenging tasks, yet it needs to be very fast in order to be clinically relevant. To make IMPT computationally tractable, robust and efficient optimization methods are required. Lastly, IMPT planning is very sensitive to accurate knowledge of relative stopping and scattering powers of the intervening tissues as well as intra- and inter-fraction motion. Robust planning methods are being developed in order to obtain IMPT plans that are less sensitive against such uncertainties. This therapy symposium will present an update on the current status and emerging developments of IMPT from the medical physics perspective. Learning Objectives: Become familiar with current delivery techniques for IMPT and their limitations. Understand the basics of dose calculational algorithms and commissioning of IMPT. Learn how to assess the accuracy of planning and delivery of IMPT treatments. Get an overview of currently used and emerging optimization techniques. Learn about the pro or contra arguments for using IMPT for moving targets.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2020

Generalized Canonical Polyadic Tensor Decomposition

Tensor decomposition is a fundamental unsupervised machine learning method in data science, with applications including network analysis and sensor data processing. This work develops a generalized canonical polyadic (GCP) low-rank tensor decomposition that allows other loss functions besides squared error. For instance, we can use logistic loss or Kullback--Leibler divergence, enabling tensor decomposition for binary or count data. We present a variety of statistically motivated loss functions for various scenarios. We provide a generalized framework for computing gradients and handling missing data that enables the use of standard optimization methods for fitting the model. Furthermore, we demonstrate the flexibility of the GCP decomposition on several real-world examples including interactions in a social network, neural activity in a mouse, and monthly rainfall measurements in India.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Surrogate Model Based Optimization for Finding Robust Deep Learning Model Architectures

Deep Learning (DL) models are increasingly used throughout the sciences. However, their performance and usefulness depend greatly on their architecture which is defined by hyperparameters such as the number of nodes, layers, the learning rate, etc. Tuning these hyperparameters is time-consuming because evaluating their performance requires a lengthy training step. Stochastic optimizers used in training lead to performance variability and potentially prediction reliability issues. In this talk, we will describe an automated optimization method based on surrogate models and active learning strategies for tuning DL model architectures. We take into account the prediction variability with the goal to identify architectures that make reliable and robust predictions. We demonstrate our developments on an application arising in particle physics.

deep learning↗
DOE OSTI2025

Power-Capping Metric Evaluation for Improving Energy Efficiency in HPC Applications

With high-performance computing systems now running at exascale, optimizing power-scaling management and resource utilization has become more critical than ever. This paper explores runtime power-capping optimizations that leverage integrated CPU-GPU power management on architectures like the NVIDIA GH200 superchip. We evaluate energy-performance metrics that account for simultaneous CPU and GPU power-capping effects by using two complementary approaches: speedup-energy-delay and a Euclidean distance-based multi-objective optimization method. By targeting a mostly compute-bound exascale science application, the Locally Self-Consistent Multiple Scattering (LSMS), we explore challenging scenarios to identify potential opportunities for energy savings in exascale applications, and we recognize that even modest reductions in energy consumption can have significant overall impacts. Our results highlight how GPU task-specific dynamic power-cap adjustments combined with integrated CPU-GPU power steering can improve the energy utilization of certain GPU tasks, thereby laying the groundwork for future adaptive optimization strategies.

Patrou, Maria [ORNL] (ORCID:0000000339754638)↗
DOE OSTI2022

A Design Method for Low-Pressure Venturi Nozzles

The purpose of this work is to provide empirical design models for low-pressure, subsonic Venturi nozzles. Experimentally validated simulations were used to determine the effect of nozzle geometry and operating conditions on the suction ratio (ratio of suction mass flow rate to motive mass flow rate) of low-pressure, subsonic Venturi nozzles, over a wide range of geometries and operating conditions, through a parametric study. The results of the parametric study were used to develop seven empirical models, each with a different range of applicability or calculating a different indicator of nozzle performance (i.e., suction ratio, momentum ratio, or dynamic pressure ratio), of the Venturi nozzles using a constrained multi-variable global optimization method. Of the seven empirical models, the best models were found to be those for low- (less than one) and high-suction ratios (greater than one), with mean absolute percentage errors of 5% and 18%, respectively. These empirical models provide a design tool for subsonic, low-pressure Venturi nozzles that is more than an order of magnitude more accurate than a governing equation approach or conventional flow head calculations. These newly-developed empirical models can be applied for initial nozzle design when precise suction ratios are required.

O’Hern, Hannah↗
DOE OSTI2024

Rapid Optimization of Total Variation with Applications in Imaging, Additive Manufacturing, and Qualification

Total Variation optimization penalizes the gradient of a control variable or state. While this work focuses on image processing in particular, it has also found applications in inverse problems and topology optimization. In image processing, the goal is to maintain faithfulness to the original image while denoising and/or deblurring. Additionally, bilevel optimization over the spatially varying regularization weights can illuminate interfaces such as damage regions and other anomalies. We will address two fundamental challenges with TV-optimization: (i) the typical slow convergence of existing TV-optimization methods, and (ii) the selection of spatially varying TV parameters to promote interface detection. Additionally, we will apply such techniques to image data collected in additive manufacturing. In said context, stochasticity in build events induces flaws in the manufactured piece, compromising the integrity of said part. There is a critical need for in-situ monitoring to spot anomalies once they form, and in this setting we apply our total variation and hyperparameter solvers. We will develop a customized algorithm based on for extreme-scale TV-optimization that achieves super-linear or quadratic-convergence, a critical property for real-time, image-by-image analysis. A worst-case outcome is a preprocessing step that enhances image quality in-situ, specifically for out-of-focus and noisy images.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Supported Single Fe Atoms Prepared via Atomic Layer Deposition for Catalytic Reactions

In this work, we report a general strategy to synthesize Fe single-atom catalysts (SACs) on various substrates (e.g., multiwalled carbon nanotubes, SiO 2 , and TiO 2 ) with high Fe loading (>1.5 wt %) by atomic layer deposition (ALD) through optimization of ferrocene (Fe precursor) dose time. Fe/SiO 2 and Fe/TiO 2 SACs were used for CO oxidation reactions and catalytic degradation of a methylene blue (MB) solution, respectively. Compared with reported iron oxide catalysts, the efficiency of Fe/SiO 2 for CO oxidation was more than 2 orders of magnitude higher than the reported values. The photocatalysis experiment demonstrated that TiO 2 nanoparticles, deposited with two cycles of Fe ALD, showed the highest activity and had a more than 6-fold photocatalytic activity enhancement over pure TiO 2 for the degradation of MB. This universal method, optimized Fe ALD, expands the potential applications of Fe SACs on different substrates.

36 MATERIALS SCIENCE↗
DOE OSTI2016

TU-AB-BRC-10: Modeling of Radiotherapy Linac Source Terms Using ARCHER Monte Carlo Code: Performance Comparison of GPU and MIC Computing Accelerators

Purpose: (1) To perform phase space (PS) based source modeling for Tomotherapy and Varian TrueBeam 6 MV Linacs, (2) to examine the accuracy and performance of the ARCHER Monte Carlo code on a heterogeneous computing platform with Many Integrated Core coprocessors (MIC, aka Xeon Phi) and GPUs, and (3) to explore the software micro-optimization methods. Methods: The patient-specific source of Tomotherapy and Varian TrueBeam Linacs was modeled using the PS approach. For the helical Tomotherapy case, the PS data were calculated in our previous study (Su et al. 2014 41(7) Medical Physics). For the single-view Varian TrueBeam case, we analytically derived them from the raw patient-independent PS data in IAEA’s database, partial geometry information of the jaw and MLC as well as the fluence map. The phantom was generated from DICOM images. The Monte Carlo simulation was performed by ARCHER-MIC and GPU codes, which were benchmarked against a modified parallel DPM code. Software micro-optimization was systematically conducted, and was focused on SIMD vectorization of tight for-loops and data prefetch, with the ultimate goal of increasing 512-bit register utilization and reducing memory access latency. Results: Dose calculation was performed for two clinical cases, a Tomotherapy-based prostate cancer treatment and a TrueBeam-based left breast treatment. ARCHER was verified against the DPM code. The statistical uncertainty of the dose to the PTV was less than 1%. Using double-precision, the total wall time of the multithreaded CPU code on a X5650 CPU was 339 seconds for the Tomotherapy case and 131 seconds for the TrueBeam, while on 3 5110P MICs it was reduced to 79 and 59 seconds, respectively. The single-precision GPU code on a K40 GPU took 45 seconds for the Tomotherapy dose calculation. Conclusion: We have extended ARCHER, the MIC and GPU-based Monte Carlo dose engine to Tomotherapy and Truebeam dose calculations.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Constitutive model development of aluminum alloy 1100 for elevated temperature forming process

Commercially pure aluminum alloy, AA1100, presents good electrical and thermal conductivity, high formability, and low cost. Those favorable characteristics have the potential to enable bipolar plates with improved economics and enhanced performance compared to current stainless steel bipolar plates for proton exchange membrane fuel cells. An accurate constitutive model is essential to develop and optimize processing parameters and effectively control the forming process. Here, the objective of this work is to develop a constitutive model of AA1100 that is able to simulate stress-strain relation, formed geometry, and predict the onset of fracture strain to avoid forming failure. Initially, a set of tensile tests at temperature between 300 and 500°C and strain rate between 0.005 and 1.0/s were conducted to examine the deformation behavior. Then, a set of damage-based unified visco-plastic constitutive equations is proposed and calibrated based on the results of stress-strain data. A genetic algorithm optimization method is applied to search for best fitting material constants in constitutive equations. The proposed model shows good predictability of both the stress-strain relation and fracture strain at low strain rate and high temperature conditions. The accuracy of proposed model is also evaluated statistically. A comparison of the proposed model with three popular models (Arrhenius-type mode, Johnson-Cook model and Zerilli-Armstrong model) was made. The proposed model shows the best experimental agreement with correlation coefficient of 0.96 in contrast to 0.25, 0.38 and 0.75 for the popular models, respectively. The proposed model can help to optimize the elevated temperature forming process and guide die design to enable optimal geometric features in the formed components.

08 HYDROGEN↗
DOE OSTI2016

Critical coupling using the hexagonal boron nitride crystals in the mid-infrared range

We theoretically demonstrate the perfect absorption phenomena in the hexagonal boron nitride (hBN) crystals in the mid-infrared wavelength ranges by means of critical coupling with a one-dimensional photonic crystal spaced by the air. Different from the polymer absorbing layer composed by a metal-dielectric composite film, the hyperbolic dispersion characteristics of hBN can meet the condition of critical coupling and achieve the total absorption in the mid-infrared wavelength ranges. However, the critical coupling phenomenon can only appear in the hBN crystals with the type II dispersion. Moreover, we discuss the influence of the thickness of hBN, the incident angle, and the thickness and permittivity of the space dielectric on the total absorption. Ultimately, the conditions for absorption enhancement and the optimization methods of perfect absorption are proposed, and the design rules for a totally absorbing system under the different conditions are achieved.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

Application of the FW-CADIS variance reduction method to calculate a precise N-flux distribution for the FRJ-2 research reactor - 15603

In order to avoid unnecessary radiation exposure and to minimize the radioactive waste and respectively cost by the decommissioning of nuclear facilities, the radiological characteristics and the activity atlases are required. Therefore, a significant step is the accurate determination of the neutron fluence distribution resulting from the operation of a reactor. For this purpose, computer codes based on the Monte-Carlo method are employed because of the flexibility in the modelling of the complex structures and physical processes. In the scope of the research project founded by the German Federal Ministry of Education and Research (BMBF), and implemented by the Institute of Nuclear Engineering and Technology Transfer (NET) -RWTH Aachen, sophisticated simulation tools are developed, which have the capability to cover the range of demands for activity and dose rate distribution using the example of the German research reactor FRJ-2. Due to the complex geometry of FRJ-2, which consists of different structures and configuration, the MCNP5 Monte Carlo Code is the optimal method for a precise neutron transport simulation. This work presents a method to achieve a precise N-Flux distribution for the FRJ-2 research reactor using hybrid deterministic Monte Carlo method. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2015

SU-E-T-597: Parameterization of the Photon Beam Dosimetry for a Commercial Linear Accelerator

Purpose: In radiation therapy, accurate data acquisition of photon beam dosimetric quantities is important for (1) beam modeling data input into a treatment planning system (TPS), (2) comparing measured and TPS modelled data, (3) a linear accelerator’s (linac) beam characteristics quality assurance process, and (4) establishing a standard data set for data comparison, etcetera. Parameterization of the photon beam dosimetry creates a portable data set that is easy to implement for different applications such as those previously mentioned. The aim of this study is to develop methods to parameterize photon percentage depth doses(PDD), profiles, and total scatter output factors(Scp). Methods: Scp, PDDs and profiles for different field sizes (from 2×2 to 40×40cm{sup 2}), depths and energies were measured in a linac using a three-dimensional water tank. All data were smoothed and profile data were also centered, symmetrized and geometrically scaled. The Scp and PDD data were analyzed using exponential functions. For modelling of open and wedge field profiles, each side was divided into three regions described by exponential, sigmoid and Gaussian equations. The model’s equations were chosen based on the physical principles described by these dosimetric quantities. The equations’ parameters were determined using a least square optimization method with the minimal amount of measured data necessary. The model’s accuracy was then evaluated via the calculation of absolute differences and distance–to–agreement analysis in low gradient and high gradient regions, respectively. Results: All differences in the PDDs’ buildup and the profiles’ penumbra regions were less than 2 mm and 0.5 mm, respectively. Differences in the low gradient regions were 0.20 ± 0.20% and 0.50 ± 0.35% for PDDs and profiles, respectively. For Scp data, all differences were less than 0.5%. Conclusion: This novel analytical model with minimum measurement requirements proved to accurately calculate PDDs, profiles, and Scp for different field sizes, depths and energies.

61 RADIATION PROTECTION AND DOSIMETRY↗