Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “algorithmic efficiency”

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 1,675 records · Page 93

A Review of Physics-Informed Machine Learning in Fluid Mechanics

Physics-informed machine-learning (PIML) enables the integration of domain knowledge with machine learning (ML) algorithms, which results in higher data efficiency and more stable predictions. This provides opportunities for augmenting—and even replacing—high-fidelity numerical simulations of complex turbulent flows, which are often expensive due to the requirement of high temporal and spatial resolution. In this review, we (i) provide an introduction and historical perspective of ML methods, in particular neural networks (NN), (ii) examine existing PIML applications to fluid mechanics problems, especially in complex high Reynolds number flows, (iii) demonstrate the utility of PIML techniques through a case study, and (iv) discuss the challenges and opportunities of developing PIML for fluid mechanics.

42 ENGINEERING↗

Extensive site-directed mutagenesis reveals interconnected functional units in the alkaline phosphatase active site

Enzymes enable life by accelerating reaction rates to biological timescales. Conventional studies have focused on identifying the residues that have a direct involvement in an enzymatic reaction, but these so-called ‘catalytic residues’ are embedded in extensive interaction networks. Although fundamental to our understanding of enzyme function, evolution, and engineering, the properties of these networks have yet to be quantitatively and systematically explored. We dissected an interaction network of five residues in the active site of Escherichia coli alkaline phosphatase. Analysis of the complex catalytic interdependence of specific residues identified three energetically independent but structurally interconnected functional units with distinct modes of cooperativity. From an evolutionary perspective, this network is orders of magnitude more probable to arise than a fully cooperative network. From a functional perspective, new catalytic insights emerge. Further, such comprehensive energetic characterization will be necessary to benchmark the algorithms required to rationally engineer highly efficient enzymes.

59 BASIC BIOLOGICAL SCIENCES↗

Lifetime Prediction of IGBT in a STATCOM Using Modified-Graphical Rainflow Counting Algorithm

Rainflow algorithms are one of the best counting methods used in fatigue and failure analysis [17]. There have been many approaches to the rainflow algorithm, some proposing modifications. Graphical Rainflow Method (GRM) was proposed recently with a claim of faster execution times [10]. However, the steps of the graphical method of rainflow algorithm, when implemented, do not generate the same output as the four-point or ASTM standard algorithm. A modified graphical method is presented and discussed in this paper to overcome the shortcomings of graphical rainflow algorithm. A fast rainflow algorithm based on four-point algorithm but considering point comparison than range comparison is also presented. A comparison between the performances of the common rainflow algorithms [6-10], including the proposed methods, in terms of execution time, memory used, and efficiency, complexity, and load sequences is presented. Finally, the rainflow algorithm is applied to temperature data of an IGBT in assessing the lifetime of a STATCOM operating for power factor correction of the load. From 5-minute data load profiles available, the lifetime is estimated to be at 3.4 years.

Gopi Reddy, Lakshmi Reddy↗

LCMSNet v. 3.0

LCMSnet communicates with, controls and orchestrates devices for high pressure liquid chromatography. In developing cutting edge chromatographic methods it was apparent that components from different manufacturers were required. Autosamplers, valves, and pumps all had to be controlled and configured to run unattended and in the highest throughput possible. The power of LCMSnet is the ability so build methods in a simple step by step fashion and then the software uses a scheduling algorithm to run the devices as efficiently as possible without crashing critical functions. This latest iteration of the software allows for modeling of the valves and incorporates logic that will report if a fluid path is undesirable. LCMSnet has reduced method development time from days to minutes and allows the ability to check configurations before any valves or pumps are connected.

Source record↗

miniSMAC2D

In the drive to make exascale computing a reality, significant effort is being invested in efficiently implementing and improving current numerical algorithms on various vendors' newest testbeds acquired by DOE's Advanced Simulation and Computing program. Porting large-scale applications to these testbeds is prohibitive in labor and time. Instead, it is more advantageous to run small-scale programs that capture the essence of the large-scale application and that can be more easily modified, improved, and distributed to other collaborators focused on hardware/software co-design. These small-scale applications are called miniapps are are being developed under the aegis of the Mantevo project.

Barnette, Daniel W.↗

FatMan vs. LittleBoy: Scaling up Linear Algebraic Operations in Scale-out Data Platforms

Linear algebraic operations such as matrix manipulations form the kernel of many machine learning and other crucial algorithms. Scaling up as well as scaling out such algorithms are highly desirable to enable efficient processing over millions of data points. To this end, we present a matrix manipulation approach to effectively scale-up each node in a scale-out data parallel platform such as Apache Spark. Specifically, we enable hardware acceleration for matrix multiplications in a distributed Spark setup without user intervention. Our approach supports both dense and sparse distributed matrices, and provides flexible control of acceleration by matrix density. We demonstrate the benefit of our approach for generalized matrix multiplication operations over large matrices with up to four billion elements. To connect the effectiveness of our approach with machine learning applications, we performed Gramian matrix computation via generalized matrix multiplications. Our experiments show that our approach achieves more than 2x performance speed-up, and up to 96.1% computation improvement, compared to a state of the art Spark MLlib for dense matrices.

Lim, Seung-Hwan↗

Storyline Visualization: A Compelling Way to Understand Patterns over Time and Space

Storyline visualization is a compelling way to understand patterns over time and space. Much effort has been spent developing efficient and aesthetically pleasing layout optimization algorithms. But what if those algorithms are optimizing the wrong things? To answer this question, we conducted a design study with different storyline layout algorithms. We found that users with our new design principles for storyline visualization outperform existing methods.

97 MATHEMATICS AND COMPUTING↗

Robust Online Monitoring for Calibration Assessment of Transmitters and Instrumentation

Robust online monitoring (OLM) technologies are expected to enable the extension or elimination of periodic sensor calibration intervals in operating and new reactors. These advances in OLM technologies will improve the safety and reliability of current and planned nuclear power systems through improved accuracy and increased reliability of sensors used to monitor key parameters. In this article, we discuss an overview of research being performed within the Nuclear Energy Enabling Technologies (NEET)/Advanced Sensors and Instrumentation (ASI) program, for the development of OLM algorithms to use sensor outputs and, in combination with other available information, 1) determine whether one or more sensors are out of calibration or failing and 2) replace a failing sensor with reliable, accurate sensor outputs. Algorithm development is focused on the following OLM functions: • Signal validation • Virtual sensing • Sensor response-time assessment These algorithms incorporate, at their base, a Gaussian Process-based uncertainty quantification (UQ) method. Various plant models (using kernel regression, GP, or hierarchical models) may be used to predict sensor responses under various plant conditions. These predicted responses can then be applied in fault detection (sensor output and response time) and in computing the correct value (virtual sensing) of a failing physical sensor. The methods being evaluated in this work can compute confidence levels along with the predicted sensor responses, and as a result, may have the potential for compensating for sensor drift in real-time (online recalibration). Evaluation was conducted using data from multiple sources (laboratory flow loops and plant data). Ongoing research in this project is focused on further evaluation of the algorithms, optimization for accuracy and computational efficiency, and integration into a suite of tools for robust OLM that are applicable to monitoring sensor calibration state in nuclear power plants.

fault detection↗

ROBUST ONLINE MONITORING FOR CALIBRATION ASSESSMENT OF TRANSMITTERS AND INSTRUMENTATION

Robust online monitoring (OLM) technologies are expected to enable the extension or elimination of periodic sensor calibration intervals in operating and new reactors. Specifically, the next generation of OLM technology is expected to include newly developed advanced algorithms that improve monitoring of sensor/system performance and enable the use of plant data to derive information that currently cannot be measured. These advances in OLM technologies will improve the safety and reliability of current and planned nuclear power systems through improved accuracy and increased reliability of sensors used to monitor key parameters. In this paper, we discuss an overview of research being performed within the Nuclear Energy Enabling Technologies (NEET)/Advanced Sensors and Instrumentation (ASI) program, for the development of OLM algorithms to use sensor outputs and, in combination with other available information, 1) determine whether one or more sensors are out of calibration or failing and 2) replace a failing sensor with reliable, accurate sensor outputs. Algorithm development is focused on the following OLM functions: • Signal validation – fault detection and selection of acceptance criteria • Virtual sensing – signal value prediction and acceptance criteria • Response-time assessment – fault detection and acceptance criteria selection A GP-based uncertainty quantification (UQ) method previously developed for UQ in OLM, was adapted for use in sensor-fault detection and virtual sensing. For signal validation, the various components to the OLM residual (which is computed using an AAKR model) were explicitly defined and modeled using a GP. Evaluation was conducted using flow loop data from multiple sources. Results using experimental data from laboratory-scale flow loops indicate that the approach, while capable of detecting sensor drift, may be incapable of discriminating between sensor drift and model inadequacy. This may be due to a simplification applied in the initial modeling, where the sensor degradation is assumed to be stationary. In the case of virtual sensors, the GP model was used in a predictive mode to estimate the correct sensor reading for sensors that may have failed. Results have indicated the viability of using this approach for virtual sensing. However, the GP model has proven to be computationally expensive, and so alternative algorithms for virtual sensing are being evaluated. Finally, automated approaches to performing noise analysis for extracting sensor response time were developed. Evaluation of this technique using laboratory-scale data indicates that it compares well with manual techniques previously used for noise analysis. Moreover, the automated and manual approaches for noise analysis also compare well with the current “gold standard”, hydraulic ramp testing, for response time monitoring. Ongoing research in this project is focused on further evaluation of the algorithms, optimization for accuracy and computational efficiency, and integration into a suite of tools for robust OLM that are applicable to monitoring sensor calibration state in nuclear power plants.

calibration interval extension↗

Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations

A method of increasing an efficiency at which a plurality of threshold gates arranged as neuromorphic hardware is able to perform a linear algebraic calculation having a dominant size of N. The computer-implemented method includes using the plurality of threshold gates to perform the linear algebraic calculation in a manner that is simultaneously efficient and at a near constant depth. “Efficient” is defined as a calculation algorithm that uses fewer of the plurality of threshold gates than a naïve algorithm. The naïve algorithm is a straightforward algorithm for solving the linear algebraic calculation. “Constant depth” is defined as an algorithm that has an execution time that is independent of a size of an input to the linear algebraic calculation. The near constant depth comprises a computing depth equal to or between O(log(log(N)) and the constant depth.

97 MATHEMATICS AND COMPUTING↗

SODA: a New Synthesis Infrastructure for Agile Hardware Design of Machine Learning Accelerators

Next generation systems, such as edge devices, will have to provide efficient processing of machine learning (ML) algorithms along several metrics, including energy, performance, area, and latency. However, the quickly evolving field of ML makes it extremely difficult to generate accelerators able to support a wide variety of algorithms. At the same time, designing accelerators in hardware description languages (HDLs) by hand is hard and time consuming, and does not allow quick exploration of the design space. This paper discusses the SODA synthesizer, an automated open source high-level ML framework-to-Verilog compiler targeting ML Application-Specific Integrated Circuits (ASICs) chiplets based on the LLVM infrastructure. The SODA synthesizers will allow implementing optimal designs by combining templated and fully tunable IPs and macros, and fully custom components generated through high-level synthesis. All these components will be provided through an extendable resource library, characterized with both commercial and open source logic design flows. Through a closed loop design space exploration engine, developers will be able to quickly explore their hardware designs along different dimension

Minutoli, Marco↗

Preparing an Incompressible-Flow Fluid Dynamics Code for Exascale-Class Wind Energy Simulations: Preprint

The US Department of Energy has identified Exascale-Class wind farm simulation tools as critical to wind energy scientific discovery. A primary objective of the Exawind project is to build high-performance, predictive Computational Fluid Dynamics tools that satisfy these modeling needs. GPU accelerators will serve as the computational thoroughbreds of next generation, Exascale-Class, platforms. Here, we report on our efforts for preparing the Exawind unstructured mesh solver, Nalu-Wind, for Exascale-Class machines. For computing at this scale, a simple port of the incompressible-flow algorithms to GPUs is not sufficient. One needs novel algorithms that are application aware, memory efficient, and optimized for latest generation GPU devices to get high-performance. The result of our efforts are unstructured mesh simulations of wind turbines that use 1/6 the compute resources of Summit supercomputer at Oak Ridge National Lab. In particular, we demonstrate a first-of-its-kind, simulation using Algebraic Multigrid solvers on over 4000 GPUs.

algebraic multigrid↗

Fast Particle-based Anomaly Detection Algorithm with Variational Autoencoder

Model-agnostic anomaly detection is one of the promising approaches in the search for new beyond the standard model physics. In this paper, we present Set-VAE, a particle-based variational autoencoder (VAE) anomaly detection algorithm. We demonstrate a 2x signal efficiency gain compared with traditional subjettiness-based jet selection. Furthermore, with an eye to the future deployment to trigger systems, we propose the CLIP-VAE, which reduces the inference-time cost of anomaly detection by using the KL-divergence loss as the anomaly score, resulting in a 2x acceleration in latency and reducing the caching requirement.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

MLOps for Beam Controls

Machine learning operations (MLOps) is the standardization and streamlining of the ML development lifecycle to address the challenges associated with large-scale machine learning applications. The full MLOps pipeline consists of open-source tools: DataHub, MinIO and MLflow. It is being used for dataset management and model development to handle changing data dependencies, varying business needs, reproducibility, and diverse teams working with differing tools and skills. To demonstrate the completion of an MLOps pipeline for particle accelerator operations, we are deploying a simple script that computes settings for the Booster’s gradient magnet power supply. Once the demonstration is complete, we will develop and deploy ML-based optimization algorithms to improve Booster’s overall efficiency. This MLOps pipeline opens the gate to systematically develop and deploy ML applications for accelerator controls and diagnostics.

43 PARTICLE ACCELERATORS↗

MLOps for Beam Controls

Machine learning operations (MLOps) is the standardization and streamlining of the ML development lifecycle to address the challenges associated with large-scale machine learning applications. The full MLOps pipeline consists of open-source tools: DataHub, MinIO and MLflow. It is being used for dataset management and model development to handle changing data dependencies, varying business needs, reproducibility, and diverse teams working with differing tools and skills. To demonstrate the completion of an MLOps pipeline for particle accelerator operations, we are deploying a simple script that computes settings for the Booster’s gradient magnet power supply. Once the demonstration is complete, we will develop and deploy ML-based optimization algorithms to improve Booster’s overall efficiency. This MLOps pipeline opens the gate to systematically develop and deploy ML applications for accelerator controls and diagnostics.

43 PARTICLE ACCELERATORS↗

Pushing sensor network computation to the edge

Sensor Networks consist of multiple devices equipped with some sensing apparatus. The devices in the network may be homogeneous or heterogeneous, yet they will coordinate in order to accomplish a pre-defined task. With the rising interest in the use of Sensor Networks in various applications, the sensor nodes need to subsist in a dynamic environment and react in a timely fashion to environmental stimuli. Unfortunately, the current paradigm in Sensor Network relies on static tasking of the nodes to support a common task; which ultimately leads to deployment of various networks to cover a common area so long as the tasks and owners of these networks differ. Straying away from this paradigm, our work introduces a framework to enable nodes to support dynamic tasking in a dynamic environment by pushing computation to the edge through FPGA-based reconfigurable nodes with increased processing power. Furthermore, we contend that the sensing apparatus available on the nodes limits the range of applications that such nodes will support. As such, reconfigurability of the nodes can yield the most efficient and responsive hardware implementation of algorithms to support common tasks of applications. The benefits of our approach are highlighted through the introduction of a target-tracking node that is reconfigurable and provides increased response time to stimuli.

47 OTHER INSTRUMENTATION↗

Hybrid and Parallel Domain-Decomposition Methods Development to Enable Monte Carlo for Reactor Analyses

This paper describes code and methods development at the Oak Ridge National Laboratory focused on enabling high-fidelity, large-scale reactor analyses with Monte Carlo (MC). Current state-of-the-art tools and methods used to perform ''real'' commercial reactor analyses have several undesirable features, the most significant of which is the non-rigorous spatial decomposition scheme. Monte Carlo methods, which allow detailed and accurate modeling of the full geometry and are considered the ''gold standard'' for radiation transport solutions, are playing an ever-increasing role in correcting and/or verifying the deterministic, multi-level spatial decomposition methodology in current practice. However, the prohibitive computational requirements associated with obtaining fully converged, system-wide solutions restrict the role of MC to benchmarking deterministic results at a limited number of state-points for a limited number of relevant quantities. The goal of this research is to change this paradigm by enabling direct use of MC for full-core reactor analyses. The most significant of the many technical challenges that must be overcome are the slow, non-uniform convergence of system-wide MC estimates and the memory requirements associated with detailed solutions throughout a reactor (problems involving hundreds of millions of different material and tally regions due to fuel irradiation, temperature distributions, and the needs associated with multi-physics code coupling). To address these challenges, our research has focused on the development and implementation of (1) a novel hybrid deterministic/MC method for determining high-precision fluxes throughout the problem space in k-eigenvalue problems and (2) an efficient MC domain-decomposition (DD) algorithm that partitions the problem phase space onto multiple processors for massively parallel systems, with statistical uncertainty estimation. The hybrid method development is based on an extension of the FW-CADIS method, which attempts to achieve uniform statistical uncertainty throughout a designated problem space. The MC DD development is being implemented in conjunction with the Denovo deterministic radiation transport package to have direct access to the 3-D, massively parallel discrete-ordinates solver (to support the hybrid method) and the associated parallel routines and structure. This paper describes the hybrid method, its implementation, and initial testing results for a realistic 2-D quarter core pressurized-water reactor model and also describes the MC DD algorithm and its implementation.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

A deep-learning based raw waveform region-of-interest finder for the liquid argon time projection chamber

The liquid argon time projection chamber (LArTPC) detector technology has an excellent capability to measure properties of low-energy neutrinos produced by the sun and supernovae and to look for exotic physics at very low energies. In order to achieve those physics goals, it is crucial to identify and reconstruct signals in the waveforms recorded on each TPC wire. In this work, we report on a novel algorithm based on a one-dimensional convolutional neural network (CNN) to look for the region-of-interest (ROI) in raw waveforms. We test this algorithm using data from the ArgoNeuT experiment in conjunction with an improved noise mitigation procedure and a more realistic data-driven noise model for simulated events. This deep-learning ROI finder shows promising performance in extracting small signals and gives an efficiency approximately twice that of the traditional algorithm in the low energy region of ~0.03–0.1 MeV. This method offers great potential to explore low-energy physics using LArTPCs.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗