Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “algorithm selection”

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 1,729 records · Page 96

Low-Cost Heliostat for High-Flux Small-Area Receivers (Final Technical Report)

This project analyzed a two-stage heliostat concept consisting of a tracking stage and a concentrating stage. The tracking stage uses mirrors mounted on a common drive that move to track the sun. The concentrating stage consists of stationary mirrors that each have a unique angle to direct rays towards a small-area, high-flux, point-focused receiver. By splitting the collection and concentrating process into two stages, multiple small, inexpensive mirrors can share a structure and be controlled by a single drive in the tracking stage. The project effort developed modeling techniques that were specifically relevant to this two-stage heliostat concept. Both field-level and unit-level models were developed. The field-level model does not explicitly consider unit-level losses which are predicted by the unit-level model and then integrated into the field-level model through a correlation referred to as an efficiency modifier. This approach is referred to as the two-model approach; the development and demonstration of this two-model approach for a multi-stage heliostat technology is a key outcome of this work. The field-level model is used to design a field that hits a specific design day power given a set of heliostat design parameters. An oversized field is simulated and then heliostat units are removed based on their annual energy production in order to generate the highest performing field. The field reduction procedure fits a smooth curve fit to annual energy production as a function of position in the field which has the effect of reducing the noise that is otherwise caused by the Monte Carlo ray tracing technique. This approach is referred to as the annual energy fit method and substantially reduces computational run time for a given field level modeling accuracy. The annual energy fit approach enables the selection of a properly sized, high-performing field using orders of magnitude fewer rays than would otherwise be possible and the development of this approach is a second key outcome of this work. These models are used within a genetic optimization algorithm in order to optimize the geometric parameters associated with a heliostat in order to achieve the lowest cost per unit of collected design day power. The cost modeling that underlies the optimization is a simple, scaling type analysis backed up by a much more detailed Design for Manufacture and Assembly (DFMA) analysis. Although the figure of merit used for optimization was not cost per mirror area, this metric is reasonable to use as a means of comparison. The optimally designed 500 kW design has a tracking mirror specific cost of $181.85/m 2 , which is significantly larger than the target value and also larger than the current state of the art. The cost of the torque-tube type linkages contributed substantially to the overall cost. Based on this observation, potentially attractive alternative design configuration utilizing a capstan type actuation system should be investigated. Finally, NREL compared the performance of the two-stage heliostat to the performance of a focused and different sized flat conventional heliostats and showed that, as expected, additional losses versus the convention heliostat caused by a worse cosine efficiency, two stages of reflection, and interstage interactions. The two-stage heliostat requires around 75% more reflective area than a flat 1x1 meter conventional heliostat (similar to a focused heliostat) and 40% more than a flat 2x2 meter conventional heliostat.

14 SOLAR ENERGY↗

Self-validated Variance-based Methods for Sensitivity Analysis of Model Outputs

Global sensitivity analysis (GSA) has the advantage over local sensitivity analysis in that GSA does not require strong model assumptions such as linearity or monotonicity. As a result, GSA methods such as those based on variance decomposition are well-suited to multi-physics models, which are often plagued by large nonlinearities. However, as with many other sampling-based methods, inadequate sample size can badly pollute the result accuracies. A natural remedy is to adaptively increase the sample size until sufficient accuracy is obtained. This paper proposes an iterative methodology comprising mechanisms for guiding sample size selection and self-assessing result accuracy. The elegant features in the the proposed methodology are the adaptive refinement strategies for stratified designs. We first apply this iterative methodology to the design of a self-validated first-order sensitivity analysis algorithm. We also extend this methodology to design a self-validated second-order sensitivity analysis algorithm based on refining replicated orthogonal array designs. Several numerical experiments are given to demonstrate the effectiveness of these methods.

99 GENERAL AND MISCELLANEOUS↗

Adaptive Protection and Validated Models to Enable Deployment of High Penetrations of Solar PV (PV-MOD)

The availability and validation of various PV models in commercial tools differ, with some models not yet thoroughly validated for advanced inverter functionalities and reliable performance under weak system conditions. Many existing models do not fully incorporate new inverter control functions, which can affect system stability. The increasing deployment of solar PV and other inverter-based resources (IBRs), including distributed energy resources (DERs), is influencing the reliable operation of protection schemes in distribution systems and microgrids. Emerging adaptive protection schemes (APS) offer new opportunities for protecting these systems during varying configurations and DER operating conditions, though their demonstration and validation remain limited. Adaptive protection schemes face similar challenges, as they are typically designed for specific configurations. There is a growing need for tools and methodologies to streamline the deployment of adaptive protection for safe and reliable DER integration. The project main objective was to develop and validate high-fidelity generic models of solar PV facilities for stability, protection, EMT, and QSTS analyses. This objective was achieved, and these models can now be integrated into commercial software tools, enabling utilities, vendors, and developers to study high-penetration PV systems more confidently. The project also demonstrated advanced applications of these models, including the design and deployment of adaptive protection schemes in high-penetration field applications and microgrids, supporting grid safety and reliability. Several milestones were reached by the end of the project. A sophisticated inverter test plan was developed, and inverters representative of the North American marketplace were selected. EPRI and NREL tested various inverters, conforming to IEEE standards. Improvements were made to existing generic models of IBR units, IBR plants, and aggregated feeders for various analyses. The first generic electromagnetic transient (EMT) model for a solar PV plant was developed, conforming to IEEE Std 2800™-2022 and validated against laboratory measurements of a 2.2 MVA large-scale battery energy storage system (BESS) inverter. That model was then used to produce reference responses illustrating examples of validated and verified IBR plant models that pass or fail tests for technical minimum capability and performance as specified in the IEEE standard. The developed, tested, and validated generic models can be used for transmission planning, stability assessments, expansion planning, and evaluating potential future IBR interconnection requirements. They can also support interconnection screens and conformity assessments of IBR plants, including solar PV. The project significantly contributed to the ongoing standardization and model-based representation and verification of IBR responses. The project further addressed challenges of common distribution protection schemes with increasing deployment of DER by developing, validating, and demonstrating adaptive protection schemes (APS) that can improve the reliable and safe integration of DER into distribution systems. New APS were designed using improved DER models for three common distribution systems: a radial feeder, a meshed network, and a microgrid. Modeling and hardware-in-the-loop (HIL) testing of the APS were conducted, successfully showing their effectiveness and selectivity. Proof-of-concept field demonstration was achieved for two APS, i.e., one on a radial feeder and another one in a microgrid. Field demonstration could not be achieved for the APS on a meshed network, primarily due apprehension of one utility partner and also due to limited access to the protective algorithms in the network protectors. Guidelines developed from the lessons learned in the project lay out the general process followed in the design, installation, and commissioning of APS for various distribution systems. Distribution utility partners’ apprehension about field demonstration of the new APS were addressed—with varying success—by taking a stepped risk-management approach of modeling of a wide range of sensitivities first, performing in-depth proof-of-concept testing in the laboratory including HIL next, and finally deliberately implementing and commissioning the actual protection equipment and algorithms into parts of—or in parallel operation to—the three real distribution systems. Future work should include pilot projects that further show the acceptable performance of the developed APS before these schemes be rolled out more widely. Inclusion of both utility and original equipment manufacturers (OEMs) in future projects could increase chances of successful field demonstration. Despite challenges in achieving the field demonstration goal of the project for all three APS, the research significantly contributed to the innovation of adaptive protection solutions for scalable and reliable DER integration into distribution systems. This project significantly enhances the understanding of the impact of using appropriate inverter models on distribution and transmission (T&D) systems. By addressing the limitations of existing generic models, the project introduces high-fidelity models for stability, protection, electromagnetic transient (EMT), and quasi-static time series (QSTS) analyses. These models, integrated into commercial software tools, enable utilities, vendors, and developers to confidently study high-penetration PV systems. The project also demonstrates advanced applications, including adaptive protection schemes (APS) for distribution systems and microgrids, ensuring grid safety and reliability. The technical effectiveness and economic feasibility of the methods are evident through the development and validation of sophisticated inverter test plans and the selection of representative inverters. Testing by EPRI and NREL on retail, commercial, and utility-scale inverters, conforming to IEEE standards, underscores the robustness of the models. Improvements to existing generic models for various analyses further enhance their validity and applicability. The project also identifies gaps in common distribution protection schemes and designed new APS using improved DER models, demonstrating their effectiveness through modeling and hardware-in-the-loop (HIL) testing. The project’s benefits to the public are manifold. By advancing the standardization and model-based representation of IBR response, it supports transmission planning, stability assessments, and future IBR interconnection requirements. The generic models can facilitate better communication between transmission planners and developers, supporting expected IBR plant capability and performance. Additionally, the development of APS for radial feeders, meshed networks, and microgrids supports the integration of distributed energy resources (DERs) into distribution systems, enhancing grid reliability and safety. The project’s emphasis on thorough testing and simplicity in design ensures practical and scalable solutions for DER integration.

14 SOLAR ENERGY↗

An Inorganic Microsphere Composite for the Selective Removal of Cesium 137 from Acidic Nuclear Waste Solutions - Parts 1 and 2

A new inorganic ion exchange composite for removing radioactive cesium from acidic waste streams has been developed. The new material consists of ammonium molybdophosphate, (NH4)3P(Mo3O10)4•3H2O (AMP), synthesized within hollow aluminosilicate microspheres (AMP-C), which are produced as a by-product from coal combustion. The selective cesium exchange capacity of this inorganic composite was evaluated in bench-scale column tests using simulated sodium bearing waste solution as a surrogate for the acidic tank waste currently stored at the Idaho National Laboratory (INL). Total cesium loading on the columns at saturation agreed very well with equilibrium values predicted from isotherm experiments performed previously. A numerical algorithm for solving the governing partial differential equations (PDE) for cesium uptake was developed using the intraparticle mass transfer coefficient obtained from previous batch kinetic experiments. Solutions to the governing equations were generated to obtain the cesium concentration at the column effluent as a function of throughput volume using the same conditions as those used for the actual column experiments. The numerical solutions of the PDE fit the column break through data quite well for all the experimental conditions in the study. The model should therefore provide a reliable prediction of column performance at larger scales. A new inorganic ion exchange composite consisting of ammonium molybdophosphate, (NH4)3P(Mo3O10)4•3H2O (AMP), synthesized within hollow aluminosilicate microspheres (AMP-C) has been developed. Two different batches of the sorbent were produced resulting in 20% and 25% AMP loading for two and three loading cycles, respectively. The selective cesium exchange capacity of this inorganic composite was evaluated using simulated sodium bearing waste solution as a surrogate for the acidic tank waste currently stored at the Idaho National Laboratory (INL). Equilibrium isotherms obtained from these experiments were very favorable for cesium uptake and indicated maximum cesium loading of approximately 9 % by weight of dry AMP. Batch kinetic experiments were also performed to obtain the necessary data to estimate the effective diffusion coefficient for cesium in the sorbent particle. These experiments resulted in effective intraparticle cesium diffusivity coefficients of 4.99 x 10-8 cm2/min and 4.72 x 10-8 cm2/min for the 20% and 25 % AMP-C material, respectively.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗

On the use of genetic algorithm to optimize industrial assets lifecycle management under safety and budget constraints

The purpose of this paper is to describe the method and tool dedicated to optimize investments planning for industrial assets. These investments may either be preventive maintenance tasks, asset enhancements or logistic investments such as spare parts purchases. The two methodological points to investigate in such an issue are: 1. The measure of the profitability of a portfolio of investments 2. The selection and planning of an optimal set of investments 3. The measure of the risk of a portfolio of investments The measure of the profitability of a set of investments in the IPOP tool is synthesised in the Net Present Value indicator. The NPV is the sum of the differences of discounted cash flows (direct costs, forced outages...) between the situations with and without a given investment. These cash flows are calculated through a pseudo-Markov reliability model representing independently the components of the industrial asset and the spare parts inventories. The component model has been widely discussed over the years but the spare part model is a new one based on some approximations that will be discussed. This model, referred as the NPV function, takes for input an investments portfolio and gives its NPV. The second issue is to optimize the NPV. If all investments were independent, this optimization would be an easy calculation, unfortunately there are two sources of dependency. The first one is introduced by the spare part model, as if components are indeed independent in their reliability model, the fact that several components use the same inventory induces a dependency. The second dependency comes from economic, technical or logistic constraints, such as a global maintenance budget limit or a safety requirement limiting the residual risk of failure of a component or group of component, making the aggregation of individual optimum not necessary feasible. The algorithm used to solve such a difficult optimization problem is a genetic algorithm. After a description of the features of the software a test case is presented showing the influence of the optimization algorithm parameters on its efficiency to find an optimal investments planning. (authors)

97 MATHEMATICS AND COMPUTING↗

SU-E-T-37: A GPU-Based Pencil Beam Algorithm for Dose Calculations in Proton Radiation Therapy

Purpose: Recent developments in radiation therapy have been focused on applications of charged particles, especially protons. Over the years several dose calculation methods have been proposed in proton therapy. A common characteristic of all these methods is their extensive computational burden. In the current study we present for the first time, to our best knowledge, a GPU-based PBA for proton dose calculations in Matlab. Methods: In the current study we employed an analytical expression for the protons depth dose distribution. The central-axis term is taken from the broad-beam central-axis depth dose in water modified by an inverse square correction while the distribution of the off-axis term was considered Gaussian. The serial code was implemented in MATLAB and was launched on a desktop with a quad core Intel Xeon X5550 at 2.67GHz with 8 GB of RAM. For the parallelization on the GPU, the parallel computing toolbox was employed and the code was launched on a GTX 770 with Kepler architecture. The performance comparison was established on the speedup factors. Results: The performance of the GPU code was evaluated for three different energies: low (50 MeV), medium (100 MeV) and high (150 MeV). Four square fields were selected for each energy, and the dose calculations were performed with both the serial and parallel codes for a homogeneous water phantom with size 300×300×300 mm3. The resolution of the PBs was set to 1.0 mm. The maximum speedup of ∼127 was achieved for the highest energy and the largest field size. Conclusion: A GPU-based PB algorithm for proton dose calculations in Matlab was presented. A maximum speedup of ∼127 was achieved. Future directions of the current work include extension of our method for dose calculation in heterogeneous phantoms.

60 APPLIED LIFE SCIENCES↗

GAtor: A First-Principles Genetic Algorithm for Molecular Crystal Structure Prediction

We present the implementation of GAtor, a massively parallel, first-principles genetic algorithm (GA) for molecular crystal structure prediction. GAtor is written in Python and currently interfaces with the FHI-aims code to perform local optimizations and energy evaluations using dispersion-inclusive density functional theory (DFT). GAtor offers a variety of fitness evaluation, selection, crossover, and mutation schemes. Breeding operators designed specifically for molecular crystals provide a balance between exploration and exploitation. Evolutionary niching is implemented in GAtor by using machine learning to cluster the dynamically updated population by structural similarity and then employing a cluster-based fitness function. Evolutionary niching promotes uniform sampling of the potential energy surface by evolving several subpopulations, which helps overcome initial pool biases and selection biases (genetic drift). The various settings offered by GAtor increase the likelihood of locating numerous low-energy minima, including those located in disconnected, hard to reach regions of the potential energy landscape. The best structures generated are re-relaxed and re ranked using a hierarchy of increasingly accurate DFT functionals and dispersion methods. GAtor is applied to a chemically diverse set of four past blind test targets, characterized by different types of intermolecular interactions. Furthermore, the experimentally observed structures and other low-energy structures are found for all four targets. In particular, for Target II, 5-cyano-3-hydroxythiophene, the top ranked putative crystal structure is a Z' = 2 structure with PT symmetry and a scaffold packing motif, which has not been reported previously.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Stimulated X-ray Raman and Absorption Spectroscopy of Iron–Sulfur Dimers

Iron–sulfur complexes play an important role in biological processes such as metabolic electron transport. A detailed understanding of the mechanism of long-range electron transfer requires knowledge of the electronic structure of the complexes, which has traditionally been challenging to obtain, either by theory or by experiment, but the situation has begun to change with advances in quantum chemical methods and intense free electron laser light sources. We compute the spectra for stimulated X-ray Raman spectroscopy (SXRS) and absorption spectroscopy of homovalent and mixed-valence [2Fe–2S] complexes, using the ab initio density matrix renormalization group algorithm. The simulated spectra show clear signatures of the theoretically predicted dense low-lying excited states within the d–d manifold. Finally, the difference in spectral intensity between the absorption-active and Raman-active states provides a potential mechanism to selectively excite states by a proper tuning of the excitation pump, to access the electronic dynamics within this manifold.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Future Atmospheric Rivers and Impacts on Precipitation: Overview of the ARTMIP Tier 2 High–Resolution Global Warming Experiment

Atmospheric rivers (ARs) are long, narrow synoptic scale weather features important for Earth’s hydrological cycle typically transporting water vapor poleward, delivering precipitation important for local climates. Understanding ARs in a warming climate is problematic because the AR response to climate change is tied to how the feature is defined. The Atmospheric River Tracking Method Intercomparison Project (ARTMIP) provides insights into this problem by comparing 16 atmospheric river detection tools (ARDTs) to a common data set consisting of high resolution climate change simulations from a global atmospheric general circulation model. ARDTs mostly show increases in frequency and intensity, but the scale of the response is largely dependent on algorithmic criteria. Across ARDTs, bulk characteristics suggest intensity and spatial footprint are inversely correlated, and most focus regions experience increases in precipitation volume coming from extreme ARs. The spread of the AR precipitation response under climate change is large and dependent on ARDT selection.

54 ENVIRONMENTAL SCIENCES↗

Local packing modulates diversity of iron pathways and cooperative behavior in eukaryotic and prokaryotic ferritins

Ferritin-like molecules show a remarkable combination of the evolutionary conserved activity of iron uptake and release that engage different pores in the conserved ferritin shell. It was hypothesized that pore selection and iron traffic depend on dynamic allostery with no conformational changes in the backbone. In this study, we detect the allosteric networks in Pseudomonas aeruginosa bacterioferritin (BfrB), bacterial ferritin (FtnA), and bullfrog M and L ferritins (Ftns) by a network-weaving algorithm (NWA) that passes threads of an allosteric network through highly correlated residues using hierarchical clustering. The residue-residue correlations are calculated in the packing-on elastic network model that introduces atom packing into the common packing-off model. Applying NWA revealed that each of the molecules has an extended allosteric network mostly buried inside the ferritin shell. The structure of the networks is consistent with experimental observations of iron transport: The allosteric networks in BfrB and FtnA connect the ferroxidase center with the 4-fold pores and B-pores, leaving the 3-fold pores unengaged. In contrast, the allosteric network directly links the 3-fold pores with the 4-fold pores in M and L Ftns. The majority of the network residues are either on the inner surface or buried inside the subunit fold or at the subunit interfaces. We hypothesize that the ferritin structures evolved in a way to limit the influence of functionally unrelated events in the cytoplasm on the allosteric network to maintain stability of the translocation mechanisms. We showed that the residue-residue correlations and the resultant long-range cooperativity depend on the ferritin shell packing, which, in turn, depends on protein sequence composition. Switching from the packing-on to the packing-off model reduces correlations by 35%–38% so that no allosteric network can be found. The influence of the side-chain packing on the allosteric networks explains the diversity in mechanisms of iron traffic suggested by experimental approaches.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Circular dichroism in photoelectron images from aligned nitric oxide molecules

We have used velocity map photoelectron imaging to study circular dichroism of the photoelectron angular distributions (PADs) of nitric oxide following two-color resonanceenhanced two-photon ionization via selected rotational levels of the A 2 Σ + , v' = 0 state. By using a circularly polarized pump beam and a counter-propagating, circularly polarized probe beam, cylindrical symmetry is preserved in the ionization process, and the images can be reconstructed using standard algorithms. The VMI set up enables individual ion rotational states to be resolved with excellent collection efficiency, rendering the measurements considerably simpler to perform than previous measurements conducted with a conventional photoelectron spectrometer. The results demonstrate that circular dichroism is observed even when cylindrical symmetry is maintained, and serve as a reminder that dichroism is a general feature of the multiphoton ionization of atoms and molecules. Furthermore, the observed PADs are in good agreement with calculations based on parameters extracted from previous experimental results obtained by using a time-offlight electron spectrometer.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

CALIBRATION OF SEMI-ANALYTIC MODELS OF GALAXY FORMATION USING PARTICLE SWARM OPTIMIZATION

We present a fast and accurate method to select an optimal set of parameters in semi-analytic models of galaxy formation and evolution (SAMs). Our approach compares the results of a model against a set of observables applying a stochastic technique called Particle Swarm Optimization (PSO), a self-learning algorithm for localizing regions of maximum likelihood in multidimensional spaces that outperforms traditional sampling methods in terms of computational cost. We apply the PSO technique to the SAG semi-analytic model combined with merger trees extracted from a standard Lambda Cold Dark Matter N-body simulation. The calibration is performed using a combination of observed galaxy properties as constraints, including the local stellar mass function and the black hole to bulge mass relation. We test the ability of the PSO algorithm to find the best set of free parameters of the model by comparing the results with those obtained using a MCMC exploration. Both methods find the same maximum likelihood region, however, the PSO method requires one order of magnitude fewer evaluations. This new approach allows a fast estimation of the best-fitting parameter set in multidimensional spaces, providing a practical tool to test the consequences of including other astrophysical processes in SAMs.

79 ASTRONOMY AND ASTROPHYSICS↗

A differentiable perturbation-based weak lensing shear estimator

Upcoming imaging surveys will use weak gravitational lensing to study the large-scale structure of the Universe, demanding sub-per cent accuracy for precise cosmic shear measurements. We present a new differentiable implementation of our perturbation-based shear estimator (fpfs), using jax, which is publicly available as part of a new suite of analytic shear algorithms called anacal. This code can analytically calibrate the shear response of any non-linear observable constructed with the fpfs shapelets and detection modes utilizing autodifferentiation (ad), generalizing the formalism to include a family of shear estimators with corrections for detection and selection biases. Using the ad capability of jax, it calculates the full Hessian matrix of the non-linear observables, which improves the previously presented second-order noise bias correction in the shear estimation. As an illustration of the power of the new anacal framework, we optimize the effective galaxy number density in the space of the generalized shear estimators using an LSST-like galaxy image simulation for the 10 yr LSST. For the generic shear estimator, the magnitude of the multiplicative bias |m| is below 3 × 10 –3 (99.7 per cent confidence interval), and the effective galaxy number density is improved by 5 per cent. We also discuss some planned future additions to the anacal software suite to extend its applicability beyond the fpfs measurements.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

PQML: Enabling the Predictive Reproducibility on NISQ Machines for Quantum ML Applications

Quantum computing represents a groundbreaking approach to high-performance computing. In recent years, quantum computers have progressed from single-qubit processors to systems boasting over 400 qubits. The presence of such a large number of qubits offers significant advantages, including enhanced computational speed—a capability beyond classical computing methods. However, the current stage of quantum computing is referred to as the noisy intermediate-scale quantum (NISQ) era. The existence of noise in this era presents challenges in testing quantum computing applications, leading to considerable variance in application results. Furthermore, the diverse noise characteristics observed across different machines exacerbate this issue, complicating the selection of the appropriate machine for application execution. In response to these challenges, we introduce our Predictive Quantum Machine Learning (PQML) tool. This tool is designed to predict outcomes when executing identical quantum machine learning applications—specifically, a critical suite of variational quantum algorithms—across various quantum computers during the NISQ era. This effort relies on data collected over a 12-month period. To the best of our knowledge, this study represents the first attempt to ensure reproducibility across quantum computers for complex circuits. Additionally, we have developed a model capable of forecasting the accuracy of quantum computers for variational quantum algorithms, with a particular emphasis on quantum machine learning as a case study.

Senapati, Priyabrata [Kent State University]↗

OpenMSI Arrayed Analysis Tools (OMAAT) v2.0

Mass spectrometry (MS) is an analytical technique that enables all chemical species within a complex mixture to be resolved and identified based on their mass:charge ratio. MS is an increasingly important tool in life science research and development because it can resolve or separate small molecules, peptides, lipids, proteins, biopolymers and fragments of all of these structures. Unfortunately, the use of MS has been limited for lack of software tools to visualize, manipulate, store and transfer MS data. Ben Bowen, Oliver Ruebel and colleagues made a huge advance on the resolution of these limitations with the invention of "OpenMSI" (#2013-004), a computational method that makes the handling of these very large files (up to 50 Tb) much easier, especially for scientists with limited background in bioinformatics or sophisticated software applications of the type needed to handle such data files. With the OpenMSI Arrayed Analysis Tool (OMAAT, #2016-188) invention, the authors take MS data analysis one step further by providing tolls for automatic sample position optimization and ion selection. Even though sample deposition technologies have improved greatly, there is always the chance of encountering outliers in a predefined roster, especially in very large sample sets. Manual curation of the data would be extremely time consuming, so the inventors have included an algorithm that automatically finds the outliers. OMAAT is written in Python with accompanying Jupyter (formerly iPython) notebook and is fully integrated with OpenMSI.

Bowen, Benjamin↗

DLIO: A DATA-CENTRIC BENCHMARK FOR DEEP LEARNING APPLICATIONS

SF-22-136 Deep learning has been shown as a successful method for various tasks, and its popularity results in numerous open-source deep learning software tools. Deep learning has been applied to a broad spectrum of scientific domains such as cosmology, particle physics, computer vision, fusion, and astrophysics. Scientists have performed a great deal of work to optimize the computational performance of deep learning frameworks. However, the same cannot be said for I/O performance. As deep learning algorithms rely on big-data volume and variety to effectively train neural networks accurately, I/O is a significant bottleneck on large-scale distributed deep learning training. DLIO, is a novel representative benchmark suite built based on the I/O profiling of the selected workloads. DLIO can be utilized to accurately emulate the I/O behavior of modern deep learning applications. Using DLIO, application developers and system software solution architects can identify potential I/O bottlenecks in their applications and guide optimizations to boost the I/O performance leading to lower training times. The storage vendor can also use DLIO as a guide for designing and optimize the storage and filesystem targeting at deep learning application.

ZHENG, HUIHUO↗

Progressive Hedging Decomposition for Solutions of Large-Scale Process Family Design Problems

In previous work, we have introduced a mathematical model for solving a discretized version of the process family design problem. This involves two sets of decision variables. One set selects which unit module designs are included in the process platform out of a candidate set of options; the other set determines which of these unit module designs are assigned to each variant. In this work, we exploit a parallelized Progressive Hedging (PH) algorithm to solve even larger scale design problems. PH is a well-known algorithm traditionally used to solve stochastic programming problems. While our problem is not a two-stage stochastic programming problem, the structure is similar, and it can be directly mapped to the PH approach, which we employ here to solve this deterministic optimization problem. We decompose our problem by process variant. We treat the platform unit module design variables as first-stage and the assignment of unit module designs to variants as second-stage, solving the problem using mpi-sppy. We demonstrate this approach on case studies of CC, water desalination, and refrigeration.

Stinchfield, Georgia↗

Photometric classification of type Ia supernovae in the SuperNova Legacy Survey with supervised learning

In the era of large astronomical surveys, photometric classification of supernovae (SNe) has become an important research field due to limited spectroscopic resources for candidate follow-up and classification. In this work, we present a method to photometrically classify type Ia supernovae based on machine learning with redshifts that are derived from the SN light-curves. This method is implemented on real data from the SNLS deferred pipeline, a purely photometric pipeline that identifies SNe Ia at high-redshifts (0.2 < z < 1.1). Our method consists of two stages: feature extraction (obtaining the SN redshift from photometry and estimating light-curve shape parameters) and machine learning classification. We study the performance of different algorithms such as Random Forest and Boosted Decision Trees. We evaluate the performance using SN simulations and real data from the first 3 years of the Supernova Legacy Survey (SNLS), which contains large spectroscopically and photometrically classified type Ia samples. Using the Area Under the Curve (AUC) metric, where perfect classification is given by 1, we find that our best-performing classifier (Extreme Gradient Boosting Decision Tree) has an AUC of 0.98.We show that it is possible to obtain a large photometrically selected type Ia SN sample with an estimated contamination of less than 5%. When applied to data from the first three years of SNLS, we obtain 529 events. We investigate the differences between classifying simulated SNe, and real SN survey data. In particular, we find that applying a thorough set of selection cuts to the SN sample is essential for good classification. This work demonstrates for the first time the feasibility of machine learning classification in a high- z SN survey with application to real SN data.

79 ASTRONOMY AND ASTROPHYSICS↗