Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Operator learning”

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 1,279 records · Page 71
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

A machine learning based approach to online electron reconstruction at CLAS12

Online reconstruction is key for monitoring purposes and real time analysis in High Energy and Nuclear Physics experiments. A necessary component of reconstruction algorithms is particle identification that combines information left by a particle passing through several detector components to identify the particle’s type. Of particular interest to electro-production Nuclear Physics experiments such as CLAS12 is electron identification which is used to trigger data recording. A machine learning approach was developed for CLAS12 to reconstruct and identify electrons by combining raw signals at the data acquisition level from several detector components. Here, this approach achieves an electron identification purity above 75% whilst retaining an efficiency close to 100%. The machine learning tools are capable of running at high rates exceeding the data acquisition rates and will allow electron reconstruction in real-time. This work enhances online analyses and monitoring and can contribute to improved triggering at CLAS12. This machine learning driven approach will also be crucial for experiments aiming to transition to streaming readout operations where online reconstruction will be a key component of the data taking paradigm.

Artificial intelligence↗
DOE OSTI2018

Scoping Study of Machine Learning Techniques for Visualization and Analysis of Multi-source Data in Nuclear Safeguards

In implementation of nuclear safeguards, many different techniques are being used to monitor operation of nuclear facilities and safeguard nuclear materials, ranging from radiation detectors, flow monitors, video surveillance, satellite imagers, digital seals to open source search and reports of onsite inspections/verifications. Each technique measures one or more unique properties related to nuclear materials or operation processes. Because these data sets have no or loose correlations, it could be beneficial to analyze the data sets together to improve the effectiveness and efficiency of safeguards processes. Advanced visualization techniques and machine-learning based multi-modality analysis could be effective tools in such integrated analysis. In this project, we will conduct a survey of existing visualization and analysis techniques for multi-source data and assess their potential values in nuclear safeguards.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Verification of Triso Fuel Burnup Using Machine Learning Algorithms

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134 Cs, 137 Cs, 154 Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2021

VERIFICATION OF TRISO FUEL BURNUP USING MACHINE LEARNING ALGORITHMS

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134Cs, 137Cs, 154Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2017

URS/WTP's Future Leader Program and AECOM's Early Career Professional Technology Exchange - 17239

URS Energy and Construction, Inc. (URS), a subsidiary of AECOM, is the Bechtel National, Inc. (BNI) teaming subcontractor on the Hanford Tank Waste Treatment and Immobilization Plant (WTP) Project. Recently, the WTP Project has begun a transition toward facility turnover and commissioning. With that transition came the hiring of staff for the startup and commissioning workforce. With many of these individuals being Early Career Professionals (ECPs), URS recognized a need to create and implement a workforce development program to give these professionals the skills to thrive in a professional environment and provide a sense of community to improve talent retention. The result of this recognition of need was the creation of URS/WTP Future Leaders. This paper will share the successes and lessons learned related to founding the URS/WTP Future Leaders. This includes establishing programs such as the Continuous Learning Workshops, Rotational Program, and implementing a mentor program. In addition, establishing a budget and obtaining support from Human Resources will be discussed. By March 2017, the ECP program will have been in effect for approximately two years. As a separate effort, many of AECOM's major waste remediation contractors have started having Early Career Professional Technology Exchanges (ECP TE) in an effort to create collaboration between sites and to establish industry contacts that will continue to be a resource for years to come. Washington River Protection Solutions, LLC; URS/WTP; Savannah River Remediation, LLC; and Nuclear Waste Partnership, LLC have conducted ECP TE, with each site hosting an event. For these events, each participant prepares a presentation and answers questions on the presentation topic. This allows ECPs to practice presentation skills, develop skills in answering questions from their peers and management, and be the expert in the room on a particular topic. Tours of the onsite facilities also are given as a way for ECPs to gain a broader understanding of the waste remediation industry and the challenges that other sites are facing. Accomplishments, lessons learned, and future plans for ECP TE will be discussed. The ECP TE events allow AECOM, as the parent company, to provide a unique learning opportunity to the next generation of engineers within the DOE nuclear field and to ensure the safe operations of these nuclear sites for years to come. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2023

Microgrid energy scheduling under uncertain extreme weather: Adaptation from parallelized reinforcement learning agents

Microgrids are useful solutions for integrating renewable energy resources and providing seamless green electricity to minimize carbon footprint. In recent years, extreme weather events happened often worldwide and caused significant economic and societal losses. Such events bring uncertainties to the microgrid energy scheduling problems and increase the challenges of microgrid operation. Traditional optimization approaches suffer from the inaccuracy of the uncertain microgrid model and the unseen events. Existing reinforcement learning (RL) - based approaches are also hampered by the limited generalization and the increasing computational burden when stochastic formulations are required to accommodate the uncertainties. This paper proposes a new parallelized reinforcement learning (PRL) method based on the probabilistic events to handle the microgrid energy uncertainties. Specifically, several local learning agents are employed to interact with pertinent microgrid environments in a distributed manner and report outcomes to the global agent, which will optimize microgrid energy resources online during extreme events. The stochastic microgrid energy optimization problem is reformulated to include all possible scenarios with probabilities. The advantage estimate functions of learning agents are designed with a backward sweep to transfer the outcomes to the value function updating process. Two simulation studies, stochastic optimization and online testing, are performed to compare with several existing RL approaches. Results substantiate that the proposed PRL method can achieve up to 20% optimization performance improvement with 4 and 28 times less computation cost than Q-learning with experience replay and multi-agent Q-learning approaches, respectively.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

Constructing a new predictive scaling formula for ITER's divertor heat-load width informed by a simulation-anchored machine learning

Understanding and predicting divertor heat-load width λq is a critically important problem for an easier and more robust operation of ITER with high fusion gain. Previous predictive simulation data for λ q using the extreme-scale edge gyrokinetic code XGC1 [S. Ku et al., Phys. Plasmas 25, 056107 (2018)] in the electrostatic limit under attached divertor plasma conditions in three major US tokamaks [C. S. Chang et al., Nucl. Fusion 57, 116023 (2017)] reproduced the Eich and Goldston attached-divertor formula results [formula #14 in T. Eich et al., Nucl. Fusion 53, 093031 (2013) and R. J. Goldston, Nucl. Fusion 52, 013009 (2012)] and furthermore predicted over six times wider λ q than the maximal Eich and Goldston formula predictions on a full-power (Q = 10) scenario ITER plasma. After adding data from further predictive simulations on a highest current JET and highest-current Alcator C-Mod, a machine learning program is used to identify a new scaling formula for λ q as a simple modification to the Eich formula #14, which reproduces the Eich scaling formula for the present tokamaks and which embraces the wide λ q XGC for the full-current Q = 10 ITER plasma. Additionally, the new formula is then successfully tested on three more ITER plasmas: two corresponding to long burning scenarios with Q = 5 and one at low plasma current to be explored in the initial phases of ITER operation. The new physics that gives rise to the wider λ q XGC is identified to be the weakly collisional, trapped-electron-mode turbulence across the magnetic separatrix, which is known to be an efficient transporter of the electron heat and mass. Electromagnetic turbulence and high-collisionality effects on the new formula are the next study topics for XGC1.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2013

Framework For Human-Automation Collaboration: Project Status Report

The Human-Automation Collaboration (HAC) research project within the Department of Energy’s Advanced Small Modular Reactor (AdvSMR) program is investigating how advanced technologies that are planned for AdvSMRs might affect the performance and the reliability of the plant from a human factors and human performance perspective. The HAC research effort investigates the consequences of allocating functions between the operators and automated systems. More specifically, the research team is addressing how to best design the collaboration between the operators and the automated systems in a manner that has the greatest positive impact on overall plant performance and reliability. As one step to accomplish this goal, the research team reviewed available information on AdvSMR designs and identified plant functions and operator tasks that might impact the design of the HAC. The identified functions and tasks will later be vetted against insights learned from the current fleet of Light Water Reactors (LWR), other industries where humans are collaborating with highly automated systems, and the results from the activities in the AdvSMR Concepts of Operations project to inform a high-level generic design model. This generic model will be used as a basis for experimental studies to study the effects of different HAC concepts as well as evaluating new human-system interface (HSI) solutions to support a well-functioning HAC. Another important part needed to accomplish the research goal is a model of the collaboration between the human operator and the automated system. The team has developed a model of HAC that defines: (1) the important design dimensions of automation that impact automation’s use by personnel and integrated human-automation performance, (2) what aspects of human cognition, behavior, and performance mediate automation’s use by personnel, and (3) when and how the above factors affect the use of automation and the overall performance of the HAC system. The HAC model is utilized to identify areas in need of additional research, which forms the basis for planned HAC research activities. By detailing the relationships between characteristics of HAC, automation, and human cognition, the HAC model identifies what aspects of the human-automation interaction are important to consider in developing automation for AdvSMR systems. Three analytical studies have been initiated to address questions related to the HAC model. These studies are 1) Models of Teamwork, 2) Standardized HAC Performance Measurement Battery, and 3) Initiators and Triggering Conditions for Adaptive Automation. Additionally, a number of field studies have been identified and are in progress, and a set of experimental studies that are on schedule to be conducted later in the project to address the main research question. This report consists of three main parts: the preliminary identification of AdvSMR functions and tasks, a description of the updates and extensions to the HAC model, and a description of ongoing and planned analytical, field, and experimental research efforts that the research team has developed to accomplish the project’s overall goal of establishing a technical basis for designing the collaboration of personnel and automation.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2021

Detector Instrumentation, Energy Reconstruction and Data Analysis in the DUNE and NOvA Neutrino Experiments

Neutrino oscillations are so far the only experimental observation beyond the standard model since its development four decades ago. The remaining fundamental questions that can be answered by long-baseline neutrino oscillations are (1) the CP violation (whether neutrinos and antineutrinos behave the same way in oscillation), and (2) the mass hierarchy ($m_3>m_{1,2}$ or $m_{1,2}>m_3$), and (3) the octant of $\theta_{23}$ ($\theta_{23}>45^\circ, =45^\circ$ or $<45^\circ$). The PI's group at UCI focuses on long-baseline neutrino experiments DUNE and NOvA, key components of DOE's Intensity Frontier program. NOvA is the current major U.S.-based long-baseline neutrino experiment aiming to solve the mass hierarchy and $\theta_{23}$ octant. It is taking $\nu$ and $\bar{\nu}$ data from the NuMI beam at Fermilab. DUNE is the next-generation flagship neutrino experiment in the U.S. which is designed to decisively determine neutrino CP violation, mass hierarchy and $\theta_{23}$ octant. DUNE is Fermilab's first priority, and has been established as an international collaboration. During the period covered by the report, the PI's group has been exceptionally productive, with leading contributions to DUNE argon-purity monitoring detectors, NOvA/DUNE deep-learning reconstruction, and NOvA oscillation and near detector (ND) analyses. The group also is actively involved in DUNE/NOvA detector operation, data production, calibration, software upgrade and maintenance, and ProtoDUNE cold electronic QA/QC. Specifically, we have made accomplishments in the following four research projects: 1) Purity monitor R\&D and data analysis for DUNE, 2) Deep-learning-based energy reconstruction at NOvA and DUNE, 3) Oscillation analysis at NOvA, and 4) $\nu$-e elastic scattering analysis at NOvA.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Black Swan Event Small-Sample Transfer Learning (BEST-L) and its Case Study on Electrical Power Prediction in COVID-19

The black swan event will usually cause a great impact on the normal operation of society. The scarcity of such events leads to a lack of relevant data and challenges in dealing with related problems. Different situations also make the traditional methods invalid. In this paper, a transfer learning framework and a convolutional neuron network are proposed to deal with the black swan small-sample events (BEST-L). Taking the COVID-19 as a typical black swan event, the BEST-L is utilized to achieve accurate mid-term load forecasting using the relationship between economy and electricity consumption. The experiment results show that the transfer learning model can effectively learn the basic knowledge about the relationship between the adopted input and output data and use a relatively small amount of data during the black swan event to improve the target areas' generalization. The approach and results can provide an effective approach to respond and react to sudden changes quickly and effectively in similar open problems.

black swan event↗
DOE OSTI2016

Daher-TLI UCSC Wash Process - Lessons Learned from Handling a Legacy Cylinder - 16506

The Daher-TLI UF{sub 6} Cylinder Service Center (UCSC) began operation in 2013 and functions as a dedicated wash, re-certification, storage and disposal facility for 122 cm uranium hexafluoride (UF{sub 6}) cylinders. The primary purpose of the facility is to provide cylinder cleaning and re--certification services to the UF{sub 6} industry for cylinders that previously contained natural or depleted UF6. The facility also has the ability to recover most of the uranium in the form of uranyl peroxide (UO{sub 4}) or triuranium octoxide (U{sub 3}O{sub 8}). Small residues of UF{sub 6} (commonly referred to as a heel) from natural cylinders are removed and treated in a patented three-step purification process that recovers U{sub 3}O{sub 8} that meets the ASTM specification. The wastewater ordinarily contains a mixture of sodium salts (77% sodium sulfate, 23% sodium nitrate) along with minute traces of uranium (1 - 2 mg/L) and fluoride (10 - 20 mg/L). Over a period of 6 months, lab analyses were performed on the filtrate from a legacy cylinder in an attempt to separate uranium and fluoride from this stream. These pilot results established near quantitative recovery of uranium (< 1 mg/L) and near complete separation of fluorides (5 - 15 mg/L). This paper follows the efforts of the Daher-TLI UCSC to process the abnormal heeled UF{sub 6} cylinder. This paper also documents the lessons learned from this experience and how those lessons were used to improve the UCSC wash process. These modifications served to recover 99.87% of the input uranium as UO{sub 4} while simultaneously creating a waste stream with only trace amounts of impurities. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2020

Deep learning model to detect various synchrophasor data anomalies

High-density synchrophasors provide valuable information for power grid situational awareness, operation and control. Unfortunately, due to factors including communication instability and hardware failure, their data quality can be greatly deteriorated by anomalies. Since the anomalies can impact the performance of the synchrophasor applications, it is of paramount significance to propose a model to detect anomalies in synchrophasor. In this study, a convolutional neural network model is established to detect and classify the anomalies in the synchrophasor measurements. Additionally, four types of anomalies observed in actual synchrophasors including erroneous patterns, random spikes, missing points and high-frequency interferences are considered in this study. The proposed model is extensively evaluated via field-collected measurements from the synchrophasor network in Jiangsu grid, China. The superior performance of the proposed model indicates the great potential of using deep learning for the detection of abnormal synchrophasor measurements.

42 ENGINEERING↗
DOE OSTI2024

Online Detection of Power Grid Anomalies via Federated Learning

Data from sensors is critical for advanced applica- tions that support efficient, reliable, and resilient electric grid operations. Historically, data from phasor measurement units (PMU) has been utilized to develop a wide variety of wide area control and protection applications suitable for power grid control centers. However, until now, most of these could not be deployed for automated operations due to a set of data corruption challenges and uncertainty in the incoming data pipeline. In this paper, we address the problem of detecting different variety of anomalies that are evident in different high- speed power grid measurements. The paper discusses a workflow for handling problems with data acquisition and highlights some of the key findings suitable for anomaly detection in a centralized and distributed environment. The effectiveness of the proposed method was demonstrated with results utilizing realistic PMU datasets

Shinkle, Matthew W.↗
DOE OSTI2012

Recovery Act: Training Curriculum Development for Building Equipment Technicians, Operators, and Commissioning Agents/Auditors

This US DOE funded project produced ten advanced energy engineering technology courses where students learn about the latest technologies and practices in the energy industry to reduce energy use in residential, commercial and industrial settings, accounting for over 50% of the world’s energy use. A geothermal and a solar thermal system were installed and commissioned as part of this project, an MATC’s cost share. An innovative intelligent lighting lab was installed, complete with course content as well as innovative academic pathway construction, laddering students from the 2-year technical college to the 4-year engineering school.

Jacobsen, Joseph↗
DOE OSTI2019

A Passive Network Cyber Threat Intelligence Framework for Legacy Critical Control Systems using Machine Learning

The importance of Internet and communication networks in our daily life and in any organization’s daily operations is well known and cannot be overstressed. A nation’s economy is fully reliant on its critical infrastructure. Energy sector is one of the 16 Critical Infrastructure Sectors identified by the Department of Homeland Security. Securing these critical infrastructure sectors is challenging but is also of utmost priority in this day of constant and persistent cyber threats. Threat is any circumstance or event that has the potential to adversely impact an agency's assets and operations. Cyber Threat Intelligence (CTI) is the process of collection, analysis, and identification of potential cyber threats to the organization. This goal of current research performed at the Savannah River National Laboratory (SRNL), Aiken, SC, is to develop a Cyber Threat Intelligence framework for gathering Threat Intelligence passively from the network traffic from and to a real or simulated Critical Control Systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

SSTDR and FDR Detection of Un-Energized and Energized Cable Anomalies Including Thermal Degradation Using Machine Learning

Historically, cables are initially qualified for nuclear power plant use for 40 years. As plants extend their operating license to 60 and 80 years, continued use of these cables must shift to a performance-based approach since it is cost prohibitive to completely replace cables that are likely still capable of performing their design function. A variety of cable tests are available and are commonly applied during outages when the cables can be taken out of service. Frequency domain reflectometry (FDR) is one of these test methods that is being more broadly accepted and used because it not only detects anomalies along the cable with a low-voltage signal that does not stress the cable insulation, but the technique also locates the anomalies. This supports follow-up local inspection and local repair or partial replacement of a damaged cable segment. Currently, FDR testing is only applied to cables that are taken out of service since the test instrument would be damaged by operational voltages.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Selecting XFEL single-particle snapshots by geometric machine learning

A promising new route for structural biology is single-particle imaging with an X-ray Free-Electron Laser (XFEL). This method has the advantage that the samples do not require crystallization and can be examined at room temperature. However, high-resolution structures can only be obtained from a sufficiently large number of diffraction patterns of individual molecules, so-called single particles. Here, we present a method that allows for efficient identification of single particles in very large XFEL datasets, operates at low signal levels, and is tolerant to background. This method uses supervised Geometric Machine Learning (GML) to extract low-dimensional feature vectors from a training dataset, fuse test datasets into the feature space of training datasets, and separate the data into binary distributions of “single particles” and “non-single particles.” As a proof of principle, we tested simulated and experimental datasets of the Coliphage PR772 virus. We created a training dataset and classified three types of test datasets: First, a noise-free simulated test dataset, which gave near perfect separation. Second, simulated test datasets that were modified to reflect different levels of photon counts and background noise. These modified datasets were used to quantify the predictive limits of our approach. Third, an experimental dataset collected at the Stanford Linear Accelerator Center. The single-particle identification for this experimental dataset was compared with previously published results and it was found that GML covers a wide photon-count range, outperforming other single-particle identification methods. Moreover, a major advantage of GML is its ability to retrieve single particles in the presence of structural variability.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

ARRA Energy Storage Demonstration Projects: Lessons Learned and Recommendations

The American Recovery and Reinvestment Act (ARRA) of 2009 (Recovery Act) provided funding for 16 energy storage demonstration projects. The projects ranged in scope from feasibility studies and technology demonstrations to full-scale, operational energy storage plants. This investment had a significant positive impact on the grid-connected energy storage industry. The goal of this report is to summarize the lessons learned from the ARRA projects, and to make recommendations for future Department of Energy (DOE) investments. Information for this report primarily came from three sources: a questionnaire and interview with each project team; DOE energy storage program peer review presentations; and DOE reports required as part of the ARRA project. Some lessons learned were common to many projects. Development of standards, codes and protocols specific to energy storage systems will mitigate uncertainty over code compliance, streamline permitting, and should be a priority (especially related to safety). Removal of regulatory barriers that preclude optimal operation of an energy storage system with multiple applications would immediately enable further deployment (e.g., FERC standards preclude the marketing department from reliability activities). Maturity of the approach to monetization varies substantially between applications with frequency regulation as an ancillary service leading other applications. Finally, developers focused on ramp mitigation and time shifting envision a reference plant. This reference plant would scale up from the current demonstration systems and would lead to the deployment of 50 MW-scale peaker plants.

25 ENERGY STORAGE↗