Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Active 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 1,027 records · Page 57

Deep reinforcement learning assisted co-optimization of Volt-VAR grid service in distribution networks

With the increasing penetration of distributed energy resources in distribution networks, Volt-VAR control and optimization (VVC/VVO) have become very important to ensure an acceptable quality of service to all customers. System operators can rely on slow-responding utility devices, including capacitor banks and on-load tap changing transformers, along with fast-responding battery and photovoltaic (PV) inverters for the VVC/VVO implementation. Because of variations in response time of these two classes of devices, and different control actions (discrete versus continuous), coordinated and optimal scheduling and operation have become of utmost importance. Here, this paper develops a look-ahead deep reinforcement learning (DRL)-based multi-objective VVO technique to improve the voltage profile of active distribution networks, decrease network and inverter power loss, and save the operational cost of the grid. It proposes a deep deterministic policy gradient (DDPG)-based approach to schedule the optimal reactive and/or active power set-points of fast-responding inverters, and a deep Q-network (DQN)-based DRL agent to schedule the discrete decisions variables of slow-responding assets. The reactive power output of PV and battery smart inverters are scheduled at 30-minute intervals and the capacitors’ commitment status is scheduled with several hour intervals. The proposed framework is validated on the modified IEEE 34-bus and 123-bus test cases with embedded PV and PV-plus-storage. To validate the efficacy of the proposed VVO, it is compared with several scenarios, including the base case without VVO, localized droop control of DERs, DDPG-only, and twin delayed DDPG (TD3) agent-based DRL techniques. The results justify the superior performance of the proposed method to improve the voltage profile, reduce network power loss, and minimize the look-ahead grid operational cost while minimizing the undesirable power losses in inverters as a result of power factor adjustments.

24 POWER TRANSMISSION AND DISTRIBUTION↗

High-throughput calculations of catalytic properties of bimetallic alloy surfaces

Abstract A comprehensive database of chemical properties on a vast set of transition metal surfaces has the potential to accelerate the discovery of novel catalytic materials for energy and industrial applications. In this data descriptor, we present such an extensive study of chemisorption properties of important adsorbates - e.g., C, O, N, H, S, CH x , OH, NH, and S H - on 2,035 bimetallic alloy surfaces in 5 different stoichiometric ratios, i.e., 0%, 25%, 50%, 75%, and 100%. To our knowledge, it is the first systematic study to compile the adsorption properties of such a well-defined, large chemical space of catalytic interest. We propose that a collection of catalytic properties of this magnitude can assist with the development of machine learning enabled surrogate models in theoretical catalysis research to design robust catalysts with high activity for challenging chemical transformations. This database is made publicly available through the platform www.Catalysis-hub.org for easy retrieval of the data for further scientific analysis.

36 MATERIALS SCIENCE↗

Automating Bayesian inference and design to quantify acoustic particle levitation

Self-propulsion of micro- and nanoparticles powered by ultrasound provides an attractive strategy for the remote manipulation of colloidal matter using biocompatible energy inputs. Quantitative understanding of particle motion and its dependence on size, shape, and composition requires accurate characterization of the acoustic field, which depends sensitively on the experimental setup. Here, we show how automated experiments based on Bayesian inference and design can accurately and efficiently characterize the acoustic field within resonant chambers used to propel acoustic nanomotors. Repeated cycles of observation, inference, and design (OID) are guided by a physical model that describes the rate at which levitating particles approach the nodal plane. Using video microscopy, we observe the relaxation of tracer particles to this plane following the application of the acoustic field. We use sequential Monte Carlo methods to infer model parameters such as the amplitude and frequency of the resonant chamber while accounting for particle-level measurement noise and population-level heterogeneity in the field. Guided by simulated outcomes, we select the optimal design for the next experiment as to maximize the information gain in the relevant parameters. We show how this iterative process serves to discriminate between competing hypotheses and efficiently converges to accurate parameter estimates using only few automated experiments. We discuss the need for model criticism to ensure the validity of the guiding model throughout automated cycles of observation, inference, and design. Furthermore, this work demonstrates how Bayesian methods can learn the parameters of nonlinear, hierarchical models used to describe video microscopy data of active colloids.

42 ENGINEERING↗

Introduction to the Special Issue on the U.S. Response to the Fukushima Accident

Provides an introduction to the May 2012 issue of Health Physics, based on a special session at the 2011 Health Physics Society (HPS) annual meeting that focused on the United States' radiological response to the Fukushima Daiichi Nuclear Power Plant accident. This introduction outlines the papers in this important issue and describes the activities of the U.S. response participants, including the U.S. Department of Energy National Nuclear Security Administration (DOE/NNSA), Department of Defense, the U.S. Nuclear Regulatory Commission (NRC) and other organizations. Observations are provided and the stage is set for the articles in this issue which document many of the activities undertaken during the Fukushima accident and which describe challenges faced and valuable lessons learned.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

TU-C-9A-01: IROC Organization and Clinical Trial Credentialing

As a response to recommendations from a report from the Institute of Medicine, NCI is reorganizing it clinical trial groups into a National Clinical Trial Network (NCTN) that consists of four adult groups (Alliance, ECOGACRIN, NRG, and SWOG) and one children’s group (COG). NRG will house CIRO, a center to promote innovative radiation therapy research and intergroup collaboration in radiation. The quality assurance groups that support clinical trials have also been restructured. ITC, OSU Imaging corelab, Philadelphia Imaging core-lab, QARC, RPC, and RTOGQA have joined together to create the Imaging and Radiation Oncology Core (IROC) Group. IROC’s mission is to provide integrated radiation oncology and diagnostic imaging quality control programs in support of the NCI’s NCTN thereby assuring high quality data for clinical trials designed to improve the clinical outcomes for cancer patients worldwide. This will be accomplished through five core services: site qualification, trial design support, credentialing, data management, case review.These changes are important for physicist participating in NCI clinical trials to understand. We will describe in detail the IROC’s activities and five core services so that as a user, the medical physicist can learn how to efficiently utilize this group. We will describe common pitfalls encountered in credentialing for current protocols and present methods to avoid them. These may include the which benchmarks are required for NSABP B-51/RTOG 1304 and how to plan them as well as tips for phantom planning. We will explain how to submit patient and phantom cases in the TRIAD system used by IROC. Learning Objectives: To understand the basic organization of IROC, its mission and five core services To learn how to use TRIAD for patient and phantom data submission To learn how to avoid common pitfalls in credentialing for current trials.

60 APPLIED LIFE SCIENCES↗

Sestrin regulation of TORC1: Is Sestrin a leucine sensor?

Sestrins are highly conserved, stress-inducible proteins that inhibit target of rapamycin complex 1 (TORC1) signaling. After their transcriptional induction, both vertebrate and invertebrate Sestrins turn on the adenosine monophosphate (AMP)–activated protein kinase (AMPK), which activates the tuberous sclerosis complex (TSC), a key inhibitor of TORC1 activation. However, Sestrin overexpression, on occasion, can result in TORC1 inhibition even in AMPK-deficient cells. This effect has been attributed to Sestrin’s ability to bind the TORC1-regulating GATOR2 protein complex, which was postulated to control trafficking of TORC1 to lysosomes. How the binding of Sestrins to GATOR2 is regulated and how it contributes to TORC1 inhibition are unknown. New findings suggest that the amino acid leucine specifically disrupts the association of Sestrin2 with GATOR2, thus explaining how leucine and related amino acids stimulate TORC1 activity. Finally, we discuss whether and how these findings fit what has already been learned about Sestrin-mediated TORC1 inhibition from genetic studies conducted in fruit flies and mammals.

60 APPLIED LIFE SCIENCES↗

Exascale Co-Design Center for Materials in Extreme Environments (ExMatEx) Annual Report — Year 2

All activities of the Exascale Co-design Center for Materials in Extreme Environments (ExMatEx) are focused on the two ultimate goals of the project: (1) demonstrating and delivering a prototype scale-bridging materials science application based upon adaptive physics refinement, and (2) identifying the requirements for the exascale ecosystem that are necessary to perform computational materials science simulations (both single- and multi-scale). During the first year of ExMatEx, our focus was on establishing how we do computational materials science, by developing an initial suite of flexible proxy applications. These “proxy apps” are the primary vehicle for the co-design process, involving assessments and tradeoff evaluations both within the ExMatEx team, and with the entire exascale ecosystem. These interactions have formed the basis of our second year activities. The set of artifacts from these co-design interactions are the lessons learned, that are used to re-express the applications and algorithms within the context of emerging architectures, programming models, and runtime systems.

36 MATERIALS SCIENCE↗

Exascale Co-Design Center for Materials in Extreme Environments (ExMatEx) Annual Report - Year 2

All activities of the Exascale Co-design Center for Materials in Extreme Environments (Ex- MatEx) are focused on the two ultimate goals of the project: (1) demonstrating and delivering a prototype scale-bridging materials science application based upon adaptive physics refinement, and (2) identifying the requirements for the exascale ecosystem that are necessary to perform computational materials science simulations (both single- and multi-scale). During the first year of ExMatEx, our focus was on establishing how we do computational materials science, by developing an initial suite of flexible proxy applications. These “proxy apps” are the primary vehicle for the co-design process, involving assessments and tradeoff evaluations both within the ExMatEx team, and with the entire exascale ecosystem. These interactions have formed the basis of our second year activities. The set of artifacts from these co-design interactions are the lessons learned, that are used to re-express the applications and algorithms within the context of emerging architectures, programming models, and runtime systems.

36 MATERIALS SCIENCE↗

Alternate Retrieval A-105 Application Assessment Report

This report summarizes the historical development of the Confined Sluicing End Effector (CSEE) technology as well scarification technology. We also summarize the systems integration and deployment of that technology in the Gunite and Associated Tanks (GAAT) at the Oak Ridge Reservation where the retrieval system successfully removed the remaining sludge and heel in the eight tanks for tank closure and disposition. We also summarize the lessons learned from the deployment at Oak Ridge. Simulant development and testing have been key activities for any waste retrieval effort as many of the physical properties of the waste are unknown; therefore, this report discusses a bounding series of simulants that helps to reduce uncertainty for operations. These simulants have been developed for potential Hanford waste retrieval deployment. The potential key attributes for retrieval, based upon information provided by Washington River Protection Services (WRPS), of single-shell tank A-105 are discussed along with the potential application of CSEE technology to the challenges presented by the unique nature of A-105. These unique features include a major tear in the bottom of the tank, which necessitates technologies other than past practice sluicing to be deployed to prevent further intrusion of waste into the vadose zone. In addition, significant in-tank hardware makes A-105 particular challenging. However, there appear to be a number of parameters that should be investigated through additional testing or review of prior testing that may demonstrate the applicability of the CSEE technology to A-105. Testing of the Waste Retrieval End Effector (WREE), which was based on the original CSEE design, to demonstrate the ability to remove sludge and hardpan from Tank 241-C-106 will be reviewed to determine relevance to the Tank A-105 retrieval. There may be new requirements for the deployment system that accesses the tank with the CSEE because the tank bottom may not be flat, requiring the ability to maneuver the CSEE to maintain the appropriate stand-off distance from the CSEE and the tank bottom. Finally, we provide a set of recommendations, a set of task descriptions, and a range of estimated costs for a demonstration of the technology by the end of FY17. These tasks are summarized as: Complete system-level functional requirements; Procure or access the capacity to perform; Generate simulant development and simulant testing; Develop CSEE functionality; Develop demonstration plan; Perform demonstration; Analyze and document testing results; Perform project management and engineering.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Neutron Tagging From Neutrino Interactions in DUNE-ND 2x2 Prototype

The Deep Underground Neutrino Experiment (DUNE) is a long-baseline neutrino oscillation experiment that aims to measure whether CP is violated in the leptonic sector (if violated, how much) and unambiguously determine the neutrino mass ordering. DUNE consists of near and far detectors that rely on liquid argon time projection chamber (LArTPC) technology to observe neutrino interactions. The near detector (ND) will be placed in Fermilab, near the neutrino source, while the far detector (FD) will be deployed in Sanford Lab, 1.5 km deep underground, which is 1300 km away from the source. LArTPCs provide excellent particle identification and calorimetry; however, detecting neutrons is challenging, as they do not leave direct ionization signals in LArTPCs. Neutrons can carry away up to 25% of the neutrino energy, introducing a significant uncertainty in DUNE measurements. The DUNE near detector (ND) features a novel modular LArTPC with pixelated charge readout, which enhances event recons truction. The modular design enables precise correlation between ionization signals and light signals in a high-rate environment, improving the identification of delayed energy depositions from neutrons in neutrino interactions. We introduce a neutron tagging technique using the 2x2 Demonstrator, a small-scale prototype of the DUNE ND LArTPC. The analysis utilizes Monte Carlo simulations and deep-learning techniques to identify neutron-induced energy depositions and reconstruct low-energy activity.

Kufatty, Georgette [Florida State U.]↗

Next generation safeguards initiative (NGSI) program plan for safeguards by design

Safeguards by Design (SBD) is defined as the incorporation of safeguards features early in the design phase of a new nuclear facility in order to avoid the need to redesign the facility at a later date, or retrofit the completed facility. Not only can SBD avoid the need for redesign or retrofit, but consideration of safeguards features early in the facility design effort can provide for a more efficient and effective safeguards design. A program has been initiated by the United States Department of Energy during the past several years to develop, demonstrate and institutionalization SBD. This plan has been developed in parallel with a similar effort at the IAEA while taking into account their achievements and future plans. The United States SBD program is focused on (1) identification of best practices that satisfy existing safeguards requirements, (2) identification of advanced concepts where best practices can be improved, and (3) institutionalizing SBD by gaining its acceptance as a global norm for the design of new nuclear facilities. SBD guidance documents are being prepared as an aid to industry for their design activities, to describe the relationship between requirements, best practices, and advanced concepts. SBD 'lessons learned' studies have been conducted to help identify the existing best practices and potential areas for improvement. Finally, acceptance as a global norm is being pursued by way of international workshops, engagement with industry and the IAEA, and setting an example by way of its use in new nuclear facilities in the United States.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

A graph neural network (GNN) approach to basin-scale river network learning: the role of physics-based connectivity and data fusion

Abstract. Rivers and river habitats around the world are under sustained pressure from human activities and the changing global environment. Our ability to quantify and manage the river states in a timely manner is critical for protecting the public safety and natural resources. In recent years, vector-based river network models have enabled modeling of large river basins at increasingly fine resolutions, but are computationally demanding. This work presents a multistage, physics-guided, graph neural network (GNN) approach for basin-scale river network learning and streamflow forecasting. During training, we train a GNN model to approximate outputs of a high-resolution vector-based river network model; we then fine-tune the pretrained GNN model with streamflow observations. We further apply a graph-based, data-fusion step to correct prediction biases. The GNN-based framework is first demonstrated over a snow-dominated watershed in the western United States. A series of experiments are performed to test different training and imputation strategies. Results show that the trained GNN model can effectively serve as a surrogate of the process-based model with high accuracy, with median Kling–Gupta efficiency (KGE) greater than 0.97. Application of the graph-based data fusion further reduces mismatch between the GNN model and observations, with as much as 50 % KGE improvement over some cross-validation gages. To improve scalability, a graph-coarsening procedure is introduced and is demonstrated over a much larger basin. Results show that graph coarsening achieves comparable prediction skills at only a fraction of training cost, thus providing important insights into the degree of physical realism needed for developing large-scale GNN-based river network models.

54 ENVIRONMENTAL SCIENCES↗

Deep reinforcement learning for predicting kinetic pathways to surface reconstruction in a ternary alloy

Abstract The majority of computational catalyst design focuses on the screening of material components and alloy composition to optimize selectivity and activity for a given reaction. However, predicting the metastability of the alloy catalyst surface at realistic operating conditions requires an extensive sampling of possible surface reconstructions and their associated kinetic pathways. We present CatGym, a deep reinforcement learning (DRL) environment for predicting the thermal surface reconstruction pathways and their associated kinetic barriers in crystalline solids under reaction conditions. The DRL agent iteratively changes the positions of atoms in the near-surface region to generate kinetic pathways to accessible local minima involving changes in the surface compositions. We showcase our agent by predicting the surface reconstruction pathways of a ternary Ni 3 Pd 3 Au 2 (111) alloy catalyst. Our results show that the DRL agent can not only explore more diverse surface compositions than the conventional minima hopping method, but also generate the kinetic surface reconstruction pathways. We further demonstrate that the kinetic pathway to a global minimum energy surface composition and its associated transition state predicted by our agent is in good agreement with the minimum energy path predicted by nudged elastic band calculations.

36 MATERIALS SCIENCE↗

A Deep Learning-Aided Workflow for Decoding the Stress Regime of Southern Nevada

The Rock Valley fault zone in southern Nevada has a notable history of seismic activity and is the site of a future direct comparison experiment of explosion and earthquake sources. This study aims to gain insight into regional tectonic processes by leveraging recent advances in seismic monitoring capabilities to elucidate the local stress regime. A crucial step in this investigation is the accurate determination of P-wave first-motion polarities, which play a vital role in resolving earthquake focal mechanisms of small earthquakes. Here, we deploy a deep learning-based method for automatic determination of first-motion polarities to vastly expand the polarity dataset beyond what has been reviewed by human analysts. By the integrating P-wave polarities with new measurements of S/P amplitude ratios, we obtain robust focal mechanism estimates for 1306 earthquakes with a local magnitude of 1 and above occurring between 2010 and 2023 in southern Nevada. We then use the focal mechanism catalog to examine the regional stress orientation, confirming an overall trans-tensional stress regime with smaller scale complexities illuminated by individual earthquake sequences. These findings demonstrate how detailed analyses of small earthquakes can provide fundamental information for understanding earthquake processes in the region and inform future experiments at the Nevada National Security Site.

58 GEOSCIENCES↗

Artificial Intelligence for Digital Security and Protections

Proper functioning of nuclear power plants relies on a mix of well-regulated human and machine-driven workflows. This regulation supports nuclear safety through a series of processes and many of the tasks that support these processes have a repetitive nature that make artificial intelligence (AI) informed by machine learning (ML) a potential aid in a variety of tasks. AI is being evaluated for activities that include inspections, fuel processing, monitoring, and other activities. The introduction of any new technology presents a potential new attack vector. In the case of AI/ML, there are many attacks that have already been discovered and over time the attacks can be expected to follow the growth pattern observed in cyber security. While future planning is necessary, current efforts need to be established now to predict the threat emergence over the next year 10 years and mitigate potential threats. Based on these observations, AI/ML will need to become trustworthy, which corresponds to techniques and procedures that emphasize AI explainability along with resilience techniques to data, algorithms, models, and systems. This kind of system robustness is the foundation for defenses against AI/ML-specific attacks. Attempting to look forward and take a broad view of capabilities provides input to research roadmaps and the ability to distill vulnerabilities into specific use cases may provide greater assistance in understanding the technology benefits while introducing new risks. The impact of current and future AI in three areas—capabilities, challenges, and recovery strategies—represents an initial attempt at balancing both.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

HD-Bind: Encoding of Molecular Structure with Low Precision, Hyperdimensional Binary Representations

Publicly available collections of drug-like molecules have grown to comprise tens of billions of compounds due to advances in combinatorial chemistry. Traditional methods for identifying "hit" molecules from a large collection of potential drug-like candidates have relied on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have the major drawback that they require exceptional computing capabilities for even relatively small collections of molecules. Hyperdimensional Computing (HDC) is a recently-proposed learning paradigm that represents data with high-dimension binary vectors; this allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas. We consider existing HDC approaches for molecular property classification and introduce two novel encodings of a commonly-used molecular representation, the extended connectivity fingerprint (ECFP). We show that HDC-based inference methods are as much as 91 times more efficient than traditional machine learning methods, and achieve an acceleration of nearly nine orders of magnitude compared to molecular docking. Our results show that HDC accelerated methods retain competitive accuracy on a number of well-studied tasks such as molecular property predictions using the MoleculeNet dataset, and bind/no-bind activity classification using the DUD-E and LIT-PCBA datasets. Our work thus motivates further investigation into molecular representation learning to develop ultraefficient pre-screening tools.

Jones, William↗

Recent Progress of JT-60SA Project toward Plasma Operation

Superconducting tokamak JT-60SA plays an essential role in fusion research and development by supporting and complementing ITER project, providing directions to the DEMO design activity and fostering next generation scientists and engineers. Since the incident of the Equilibrium Field coil #1 during the Integrated Commissioning (IC) in March 2021, both EU and JA Implementing agencies (IAs) have examined how to ensure safety operation of JT-60SA by mitigating the risk of possible discharge occurrence inside the cryostat. Based on the experience of the Global Paschen tests, the IAs have established a strategy of risk mitigation measures, which is a combination of (i) reinforcement of insulation, (ii) avoiding unnecessary voltage application to the coil systems and (iii) immediate de-energization of the coils when deteriorated vacuum condition is detected. Thanks to the considerable efforts of the Integrated Project Team (IPT) members, the IC restarted in May 2023. After the confirmation of superconducting state of coil systems (TF, EF and CS), the coil energization test and the plasma operation (OP-1) starts. The first plasma was successfully achieved on 23 October 2023 with a limited value of applied voltage and current to the coils. The plasma configuration control will be also confirmed with low plasma current and low auxiliary heating power conditions. Based on the IO-F4E-QST collaboration, activities of JT-60SA have been shared with the IO and provided an important lesson learned for ITER assembly and commissioning, and will provide an outstanding contribution to fusion research at large. After OP-1, Maintenance & Enhancement phase 1 (M/E-1) starts from January 2024, in which in-vessel components are installed, and heating system and diagnostic system are extensively upgraded to allow high power heating experiment planned in OP-2. In order to make the best use of JT-60SA, newly organized JT-60SA experiment team will refine the research plan in the future high heating power operation phase.

Broader Approach activities↗