Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Software algorithms”

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 1,693 records · Page 94

Performance Characterization for Fusion Co-design Applications

Magnetic fusion is a long-term solution for producing electrical power for the world, and the large thermonuclear international device (ITER) being constructed will produce net energy and a path to fusion energy provided the computer modeling is accurate. To effectively address the requirements of the high-end fusion simulation community, application developers, algorithm designers, and hardware architects must have reliable simulation data gathered at scale for scientifically valid configurations. This paper presents detailed benchmarking results for a set of magnetic fusion applications with a wide variety of underlying mathematical models including both particle-in-cell and Eulerian codes using both implicit and explicit numerical solvers. Our evaluation on a petascale Cray XE6 platform focuses on profiling these simulations at scale identifying critical performance characteristics, including scalability, memory/network bandwidth limitations, and communication overhead. Overall results are a key in improving fusion code design, and are a critical first step towards exascale hardware-software co-design — a process that tightly couples applications, algorithms, imple- mentation, and computer architecture.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Reconstruction of Charged Particle Tracks in Realistic Detector Geometry Using a Vectorized and Parallelized Kalman Filter Algorithm

One of the most computationally challenging problems expected for the High-Luminosity Large Hadron Collider (HL-LHC) is finding and fitting particle tracks during event reconstruction. Algorithms used at the LHC today rely on Kalman filtering, which builds physical trajectories incrementally while incorporating material e ects and error estimation. Recognizing the need for faster computational throughput, we have adapted Kalman-filterbased methods for highly parallel, many-core SIMD and SIMT architectures that are now prevalent in high-performance hardware. Previously we observed significant parallel speedups, with physics performance comparable to CMS standard tracking, on Intel Xeon, Intel Xeon Phi, and (to a limited extent) NVIDIA GPUs. While early tests were based on artificial events occurring inside an idealized barrel detector, we showed subsequently that our mkFit software builds tracks successfully from complex simulated events (including detector pileup) occurring inside a geometrically accurate representation of the CMS-2017 tracker. Here, we report on advances in both the computational and physics performance of mkFit, as well as progress toward integration with CMS production software. Recently we have improved the overall eciency of the algorithm by preserving short track candidates at a relatively early stage rather than attempting to extend them over many layers. Moreover, mkFit formerly produced an excess of duplicate tracks; these are now explicitly removed in an additional processing step. We demonstrate that with these enhancements, mkFit becomes a suitable choice for the first iteration of CMS tracking, and eventually for later iterations as well. We plan to test this capability in the CMS High Level Trigger during Run 3 of the LHC, with an ultimate goal of using it in both the CMS HLT and oine reconstruction for the HL-LHC CMS tracker.

Cerati, Giuseppe↗

TH-A-19A-01: An Open Source Software for Proton Treatment Planning in Heterogeneous Medium

Purpose: Due to its success in Radiation Oncology during the last decade, interest in proton therapy is on the rise. Unfortunately, despite the global enthusiasm in the field, there is presently no free, multiplatform and customizable Treatment Planning System (TPS) providing proton dose distributions in heterogenous medium. This restricts substantially the progress of clinical research for groups without access to a commercial Proton TPS. The latest implementation of our pencil beam dose calculation algorithm for proton beams within the 3D Slicer open-source environment fulfills all the conditions described above. Methods: The core dose calculation algorithm is based on the Hong algorithm (1), which was upgraded with the Kanematsu theory describing the evolution of the lateral scattering of proton beamlets in heterogeneous medium. This algorithm deals with both mono-energetic beams and Spread Out Bragg Peak (SOBP). In order to be user-friendly, we provide a graphical user interface implemented with the Qt libraries, and visualization with the 3D Slicer medical image analysis software. Two different pencil beam algorithms were developed, and the clinical proton beam line at our facility was modeled. Results: The dose distributions provided by our algorithms were compared to dose distributions coming from both commercialized XiO TPS and literature (dose measurements, GEANT4 and MCNPx) and turned out to be in a good agreement, with maximum dose discrepancies of 5% in homogeneous phantoms and 10% in heterogeneous phantoms. The algorithm of SOBP creation from an optimized weigthing of mono-energetic beams results in flat SOBP. Conclusion: We hope that our efforts in implementing this new, open-source proton TPS will help the research groups to have a free access to a useful, reliable proton dose calculation software.(1) L. Hong et al., A pencil beam algorithm for proton dose calculations, Phys. Med. Biol. 41 (1996) 1305–1330. This project is paid for by NCI/MGH Federal share as a research grant.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Streamlining GNC Architecture Development and FSW Integration forthe Mars Ascent Vehicle

The Mars Ascent Vehicle (MAV) will be the first vehicle to perform an ascent from the surface ofanother atmospheric planetary body outside of the Earth-Moon system. Significant light-time delayrequires complete autonomy of flight throughout ascent, and naturally a high level of reliability isdesired in both MAV’s hardware and software subsystems. The MAV Guidance, Navigation and Controls(GNC) team and the MAV Flight Software (FSW) team have partnered together to improve the efficiencyof algorithm integration onto the MAV flight processor, and to increase confidence that said integrationis successful and without human error. An interface architecture is proposed for the GNC suite thatallows both the guidance and navigation subsystems to provide code algorithms directly in C++, and thecontrols subsystem to provide MATLAB Simulink auto-coded algorithms. Several continuous integration/deployment (CI/CD) methodologies have been considered for ease of transition of algorithm code fromthe GNC team to the FSW team. The GNC/FSW teams also worked together to develop a cFS-friendlywrapper which abstracts the integration of the GNC algorithm code into an interface-level API that iscompatible with cFS. Several iterations of vehicle GNC code have been produced between the GNC/FSWteam’s partnership, and this strong interface between these two teams have allowed the GNC/FSWteams to greatly increase confidence of efficient and error-free implementation of the GNC code ontoMAV for a successful flight.

GNC↗

Solution of monotone complementarity and general convex programming problems using a modified potential reduction interior point method

We present a homogeneous algorithm equipped with a modified potential function for the monotone complementarity problem. We show that this potential function is reduced by at least a constant amount if a scaled Lipschitz condition (SLC) is satisfied. A practical algorithm based on this potential function is implemented in a software package named iOptimize. The implementation in iOptimize maintains global linear and polynomial time convergence properties, while achieving practical performance. It either successfully solves the problem, or concludes that the SLC is not satisfied. When compared with the mature software package MOSEK (barrier solver version 6.0.0.106), iOptimize solves convex quadratic programming problems, convex quadratically constrained quadratic programming problems, and general convex programming problems in fewer iterations. Moreover, several problems for which MOSEK fails are solved to optimality. In addition, we also find that iOptimize detects infeasibility more reliably than the general nonlinear solvers Ipopt (version 3.9.2) and Knitro (version 8.0).

97 MATHEMATICS AND COMPUTING↗

Wireless Sensor Network for Advanced Energy Management Solutions

Eaton has developed an advanced energy management solution that has been deployed to several Industries of the Future (IoF) sites. This demonstrated energy savings and reduced unscheduled downtime through an improved means for performing predictive diagnostics and energy efficiency estimation. Eaton has developed a suite of online, continuous, and inferential algorithms that utilize motor current signature analysis (MCSA) and motor power signature analysis (MPSA) techniques to detect and predict the health condition and energy usage condition of motors and their connect loads. Eaton has also developed a hardware and software platform that provided a means to develop and test these advanced algorithms in the field. Results from lab validation and field trials have demonstrated that the developed advanced algorithms are able to detect motor and load inefficiency and performance degradation. Eaton investigated the performance of Wireless Sensor Networks (WSN) within various industrial facilities to understand concerns about topology and environmental conditions that have precluded broad adoption by the industry to date. A Wireless Link Assessment System (WLAS), was used to validate wireless performance under a variety of conditions. Results demonstrated that wireless networks can provide adequate performance in most facilities when properly specified and deployed. Customers from various IoF expressed interest in applying wireless more broadly for selected applications, but continue to prefer utilizing existing, wired field bus networks for most sensor based applications that will tie into their existing Computerized Motor Maintenance Systems (CMMS). As a result, wireless technology was de-emphasized within the project, and a greater focus placed on energy efficiency/predictive diagnostics. Commercially available wireless networks were only utilized in field test sites to facilitate collection of motor wellness information, and no wireless sensor network products were developed under this project. As an outgrowth of this program, Eaton developed a patented energy-optimizing drive control technology that is complementary to a traditional variable frequency drives (VFD) to enable significant energy savings for motors with variable torque applications, such as fans, pumps, and compressors. This technology provides an estimated energy saving of 2%-10% depending on the loading condition, in addition to the savings obtained from a traditional VFD. The combination of a VFD with the enhanced energy-optimizing controls will provide significant energy savings (10% to 70% depending on the load and duty cycle) for motors that are presently connected with across the line starters. It will also provide a more favorable return on investment (ROI), thus encouraging industries to adopt VFDs for more motors within their facilities. The patented technology is based on nonintrusive algorithms that estimate the instantaneous operating efficiency and motor speed and provide active energy-optimizing control of a motor, using only existing voltage and current sensors. This technology is currently being commercialized by Eaton’s Industrial Controls Division in their next generation motor control products. Due to the common nonintrusive and inferential nature of various algorithms, this same product can also include motor and equipment condition monitoring features, providing the facility owner additional information to improve process uptime and the associated energy savings. Calculations estimated potential energy savings of 261,397GWh/Yr ($15.7B/yr), through retrofitting energy-optimizing VFDs into existing facilities, and incorporating the solution into building equipment sold by original equipment manufacturers (OEMs) and installed by mechanical and electrical contractors. Utilizing MCSA and MPSA for predictive maintenance (PM) of motors and connected equipment reduces process downtime cost and the cost of wasted energy associated with shutting down and restarting the processes. Estimated savings vary depending on the industry segment and equipment criticality per facility/process. Average downtime for an industrial facility is 4-12 hours with a cost/hr of $7500/hr, with large, critical processes reaching $50-100k/hr. Specific downtime costs are not included in this report because of customer confidentiality, but projected savings across the Industries of the Future (IoF) are still expected to be comparable to the original program estimates. Two generations of customer field deployments and evaluation have been completed during the course of this project. Results from these customer sites have been used for identifying the scope and improving the developed energy and wellness algorithms. The field deployments have confirmed that the hardware for sensing and sampling motor currents and voltages are reliable and able to provide an adequate signal-to-noise ratio from the electrical noise present on the motor signals.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

An order (n) algorithm for the dynamics simulation of robotic systems

The formulation of an Order (n) algorithm for DISCOS (Dynamics Interaction Simulation of Controls and Structures), which is an industry-standard software package for simulation and analysis of flexible multibody systems is presented. For systems involving many bodies, the new Order (n) version of DISCOS is much faster than the current version. Results of the experimental validation of the dynamics software are also presented. The experiment is carried out on a seven-joint robot arm at NASA's Goddard Space Flight Center. The algorithm used in the current version of DISCOS requires the inverse of a matrix whose dimension is equal to the number of constraints in the system. Generally, the number of constraints in a system is roughly proportional to the number of bodies in the system, and matrix inversion requires O(p exp 3) operations, where p is the dimension of the matrix. The current version of DISCOS is therefore considered an Order (n exp 3) algorithm. In contrast, the Order (n) algorithm requires inversion of matrices which are small, and the number of matrices to be inverted increases only linearly with the number of bodies. The newly-developed Order (n) DISCOS is currently capable of handling chain and tree topologies as well as multiple closed loops. Continuing development will extend the capability of the software to deal with typical robotics applications such as put-and-place, multi-arm hand-off and surface sliding.

Chun, H. M.↗

Automating the design of scientific computing software

SINAPSE is a domain-specific software design system that generates code from specifications of equations and algorithm methods. This paper describes the system's design techniques (planning in a space of knowledge-based refinement and optimization rules), user interaction style (user has option to control decision making), and representation of knowledge (rules and objects). It also summarizes how the system knowledge has evolved over time and suggests some issues in building software design systems to facilitate reuse.

Kant, Elaine↗

NE-COST plug-in: Expanding ACCERT's Capabilities for Life-Cycle Cost Modeling

The Algorithm for the Capital Cost Estimation of Reactor Technologies (ACCERT) is a structured methodology and software tool designed to simplify and standardize cost estimation for nuclear reactor technologies [1]. By utilizing a relational database structure and modular cost estimation algorithms, ACCERT delivers a robust, flexible, and scalable framework for evaluating costs across various reactor types and configurations [2]. The recent integration of the NE-COST plugin further expands ACCERT’s scope by introducing detailed life-cycle cost modeling and probabilistic analysis of uncertainties. This addition enables users to evaluate costs across front-end processes such as uranium enrichment and fabrication, as well as back-end activities including waste disposal and geologic storage. Through Monte Carlo statistical cost simulations, the plugin provides probabilistic insights into cost ranges, offering critical decision-making support for stakeholders including reactor developers, policymakers, and researchers.

Zhou, Jia↗

Development and Validation of a Polar Cloud Algorithm for CERES

The objectives of this project, as described in the original proposal, were to develop an algorithm for diagnosing cloud properties over snow- and ice-covered surfaces, particularly at night, using satellite radiances from the Advanced Very High Resolution Radiometer (AVHRR) and High-resolution Infrared Radiation Sounder (HIRS) sensors. Products from this algorithm include a cloud mask and additional cloud properties such as cloud phase, amount, and height. The SIVIS software package, developed as a part of the CERES project, was originally the primary tool used to develop the algorithm, but as it is no longer supported we have had to pursue a new tool to enable the combination and analysis of collocated radiances from AVHRR and HIRS. This turned out to be a much larger endeavor than we expected, but we now have the data sets collocated (with many thanks to B. Baum for the fundamental code) and we have developed a nighttime cloud detection algorithm. Using this algorithm we have also computed realistic-looking cloud fractions from AVHRR brightness temperatures. A method to identify cloud phase has also been implemented. Atmospheric information from the TIROS Operational Vertical Sounder (TOVS) Polar Pathfinder Data Set, which includes temperature and moisture profiles as well as surface information, provides information required for determining cloud-top height.

Source record↗

Quantifying Contaminant Mass for the Feasibility Study of the DuPont Chambers Works FUSRAP Site - 13510

The U.S. Army Corps of Engineers (USACE) - Philadelphia District is conducting an environmental restoration at the DuPont Chambers Works in Deepwater, New Jersey under the Formerly Utilized Sites Remedial Action Program (FUSRAP). Discrete locations are contaminated with natural uranium, thorium-230 and radium-226. The USACE is proposing a preferred remedial alternative consisting of excavation and offsite disposal to address soil contamination followed by monitored natural attenuation to address residual groundwater contamination. Methods were developed to quantify the error associated with contaminant volume estimates and use mass balance calculations of the uranium plume to estimate the removal efficiency of the proposed alternative. During the remedial investigation, the USACE collected approximately 500 soil samples at various depths. As the first step of contaminant mass estimation, soil analytical data was segmented into several depth intervals. Second, using contouring software, analytical data for each depth interval was contoured to determine lateral extent of contamination. Six different contouring algorithms were used to generate alternative interpretations of the lateral extent of the soil contamination. Finally, geographical information system software was used to produce a three dimensional model in order to present both lateral and vertical extent of the soil contamination and to estimate the volume of impacted soil for each depth interval. The average soil volume from all six contouring methods was used to determine the estimated volume of impacted soil. This method also allowed an estimate of a standard deviation of the waste volume estimate. It was determined that the margin of error for the method was plus or minus 17% of the waste volume, which is within the acceptable construction contingency for cost estimation. USACE collected approximately 190 groundwater samples from 40 monitor wells. It is expected that excavation and disposal of contaminated soil will remove the contaminant source zone and significantly reduce contaminant concentrations in groundwater. To test this assumption, a mass balance evaluation was performed to estimate the amount of dissolved uranium that would remain in the groundwater after completion of soil excavation. As part of this evaluation, average groundwater concentrations for the pre-excavation and post-excavation aquifer plume area were calculated to determine the percentage of plume removed during excavation activities. In addition, the volume of the plume removed during excavation dewatering was estimated. The results of the evaluation show that approximately 98% of the aqueous uranium would be removed during the excavation phase. The USACE expects that residual levels of contamination will remain in groundwater after excavation of soil but at levels well suited for the selection of excavation combined with monitored natural attenuation as a preferred alternative. (authors)

54 ENVIRONMENTAL SCIENCES↗

VTK-m User's' Guide (V.1.7)

High-performance computing relies on ever finer threading. Advances in processor technology include ever greater numbers of cores, hyperthreading, accelerators with integrated blocks of cores, and special vectorized instructions, all of which require more software parallelism to achieve peak performance. Traditional visualization solutions cannot support this extreme level of concurrency. Extreme scale systems require a new programming model and a fundamental change in how we design algorithms. To address these issues we created VTK-m: the visualization toolkit for multi-/many-core architectures. VTK-m supports a number of algorithms and the ability to design further algorithms through a top-down design with an emphasis on extreme parallelism. VTK-m also provides support for finding and building links across topologies, making it possible to perform operations that determine manifold surfaces, interpolate generated values, and find adjacencies. Although VTK-m provides a simplified high-level interface for programming, its template-based code removes the overhead of abstraction. VTK-m simplifies the development of parallel scientific visualization algorithms by providing a framework of supporting functionality that allows developers to focus on visualization operations. Consider the listings in Figure 1.1 that compares the size of the implementation for the Marching Cubes algorithm in VTK-m with the equivalent reference implementation in the CUDA software development kit. Because VTK-m internally manages the parallel distribution of work and data, the VTK-m implementation is shorter and easier to maintain. Additionally, VTK-m provides data abstractions not provided by other libraries that make code written in VTK-m more versatile.This book includes contributions from the VTK-m community including the VTK-m development team and the user community.

97 MATHEMATICS AND COMPUTING↗

The VTK-m Users' Guide (V.1.9)

High-performance computing relies on ever finer threading. Advances in processor technology include ever greater numbers of cores, hyperthreading, accelerators with integrated blocks of cores, and special vectorized instructions, all of which require more software parallelism to achieve peak performance. Traditional visualization solutions cannot support this extreme level of concurrency. Extreme scale systems require a new programming model and a fundamental change in how we design algorithms. To address these issues we created VTK-m: the visualization toolkit for multi-/many-core architectures. VTK-m supports a number of algorithms and the ability to design further algorithms through a top-down design with an emphasis on extreme parallelism. VTK-m also provides support for finding and building links across topologies, making it possible to perform operations that determine manifold surfaces, interpolate generated values, and find adjacencies. Although VTK-m provides a simplified high-level interface for programming, its template-based code removes the overhead of abstraction. VTK-m simplifies the development of parallel scientific visualization algorithms by providing a framework of supporting functionality that allows developers to focus on visualization operations. Consider the listings in Figure 1.1 that compares the size of the implementation for the Marching Cubes algorithm in VTK-m with the equivalent reference implementation in the CUDA software development kit. Because VTK-m internally manages the parallel distribution of work and data, the VTK-m implementation is shorter and easier to maintain. Additionally, VTK-m provides data abstractions not provided by other libraries that make code written in VTK-m more versatile.

97 MATHEMATICS AND COMPUTING↗

The VTK-m Users' Guide (V.2.0)

High-performance computing relies on ever finer threading. Advances in processor technology include ever greater numbers of cores, hyperthreading, accelerators with integrated blocks of cores, and special vectorized instructions, all of which require more software parallelism to achieve peak performance. Traditional visualization solutions cannot support this extreme level of concurrency. Extreme scale systems require a new programming model and a fundamental change in how we design algorithms. To address these issues we created VTK-m: the visualization toolkit for multi-/many-core architectures. VTK-m supports a number of algorithms and the ability to design further algorithms through a top-down design with an emphasis on extreme parallelism. VTK-m also provides support for finding and building links across topologies, making it possible to perform operations that determine manifold surfaces, interpolate generated values, and find adjacencies. Although VTK-m provides a simplified high-level interface for programming, its template-based code removes the overhead of abstraction. VTK-m simplifies the development of parallel scientific visualization algorithms by providing a framework of supporting functionality that allows developers to focus on visualization operations. Consider the listings in Figure 1.1 that compares the size of the implementation for the Marching Cubes algorithm in VTK-m with the equivalent reference implementation in the CUDA software development kit. Because VTK-m internally manages the parallel distribution of work and data, the VTK-m implementation is shorter and easier to maintain. Additionally, VTK-m provides data abstractions not provided by other libraries that make code written in VTK-m more versatile.

97 MATHEMATICS AND COMPUTING↗

Systematic Benchmarking of Diagnostic Technologies for an Electrical Power System

Automated health management is a critical functionality for complex aerospace systems. A wide variety of diagnostic algorithms have been developed to address this technical challenge. Unfortunately, the lack of support to perform large-scale V&V (verification and validation) of diagnostic technologies continues to create barriers to effective development and deployment of such algorithms for aerospace vehicles. In this paper, we describe a formal framework developed for benchmarking of diagnostic technologies. The diagnosed system is the Advanced Diagnostics and Prognostics Testbed (ADAPT), a real-world electrical power system (EPS), developed and maintained at the NASA Ames Research Center. The benchmarking approach provides a systematic, empirical basis to the testing of diagnostic software and is used to provide performance assessment for different diagnostic algorithms.

Kurtoglu, Tolga↗

Privacy Preserving Federated Learning for Advanced Scientific Ecosystems

We present a framework to provide privacy preserving (PP) federating learning (FL) across multiple computational and experimental facilities. This work joins the compute capabilities of National Energy Research Scientific Computing Center (NERSC) and Oak Ridge National Laboratory Research Cloud (ORC) with simulated experimental data, such as those produced at the SLAC National Accelerator Laboratory and Spallation Neutron Source (SNS). We describe the software infrastructure developed to provide privacy for computational and experimental networks. We developed algorithmic privacy across the federated system by embedding database security, computation, and communication into the federation architecture, utilizing scientific tools developed by the experimental community.

Archibald, Rick [ORNL] (ORCID:0000000245389780)↗

Productive Programming of Distributed Systems with the SHAD C++ Library

High-performance computing (HPC) is often perceived as a matter of making large-scale systems (e.g., clusters) run as fast as possible, regardless the required programming effort. However, the idea of "bringing HPC to the masses" has recently emerged. Inspired by this vision, we have designed SHAD, the Scalable High-performance Algorithms and Data-structures library. SHAD is open source software, written in C++, for C++ developers. Unlike other HPC libraries for distributed systems, which rely on SPMD models, SHAD adopts a shared-memory programming abstraction, to make C++ programmers feel at home. Underneath, SHAD manages tasking and data-movements, moving the computation where data resides and taking advantage of asynchrony to tolerate network latency. At the bottom of his stack, SHAD can interface with multiple runtime systems: this not only improves developer’s productivity, by hiding the complexity of such software and of the underlying hardware, but also greatly enhance code portability. Thanks to its abstraction layers, SHAD can indeed target different systems, ranging from laptops to HPC clusters, without any need for modifying the user-level code. We have prototyped and open-sourced the implementation of (a subset of) the C++ standard library (STL) targeting multi-node HPC clusters. Our work allows plain STL-based C++ code to scale on HPC systems, with no need for rewriting the code to exploit the complex hardware. SHAD is available under Apache v2 License at https://github.com/pnnl/SHAD. In this paper we overview the design of the SHAD library, depicting its main components: runtime systems abstractions for tasking; parallel and distributed data-structures; STL-compliant interfaces and algorithms.

Castellana, Vito G.↗