Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “adaptive 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 1,189 records · Page 66

Reconfigurable perovskite nickelate electronics for artificial intelligence

Reconfigurable devices offer the ability to program electronic circuits on demand. Here, in this work, we demonstrated on-demand creation of artificial neurons, synapses, and memory capacitors in post-fabricated perovskite NdNiO 3 devices that can be simply reconfigured for a specific purpose by single-shot electric pulses. The sensitivity of electronic properties of perovskite nickelates to the local distribution of hydrogen ions enabled these results. With experimental data from our memory capacitors, simulation results of a reservoir computing framework showed excellent performance for tasks such as digit recognition and classification of electrocardiogram heartbeat activity. Using our reconfigurable artificial neurons and synapses, simulated dynamic networks outperformed static networks for incremental learning scenarios. The ability to fashion the building blocks of brain-inspired computers on demand opens up new directions in adaptive networks.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Lorentz group equivariant autoencoders

Abstract There has been significant work recently in developing machine learning (ML) models in high energy physics (HEP) for tasks such as classification, simulation, and anomaly detection. Often these models are adapted from those designed for datasets in computer vision or natural language processing, which lack inductive biases suited to HEP data, such as equivariance to its inherent symmetries. Such biases have been shown to make models more performant and interpretable, and reduce the amount of training data needed. To that end, we develop the Lorentz group autoencoder (LGAE), an autoencoder model equivariant with respect to the proper, orthochronous Lorentz group $$\textrm{SO}^+(3,1)$$ SO + ( 3 , 1 ) , with a latent space living in the representations of the group. We present our architecture and several experimental results on jets at the LHC and find it outperforms graph and convolutional neural network baseline models on several compression, reconstruction, and anomaly detection metrics. We also demonstrate the advantage of such an equivariant model in analyzing the latent space of the autoencoder, which can improve the explainability of potential anomalies discovered by such ML models.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Design and construction of coal/biomass to liquids (CBTL) process development unit (PDU) at the University of Kentucky Center for Applied Energy Research (CAER)

This report describes a first phase of a project to design, construct and commission an integrated coal/biomass-to-liquids facility at a capacity of 1 bbl. /day at the University of Kentucky Center for Applied Energy Research (UK-CAER) – specifically for construction of the building and upstream process units for feed handling, gasification, and gas cleaning, conditioning and compression. The deliverables from the operation of this pilot plant [when fully equipped with the downstream process units] will be firstly the liquid FT products and finished fuels which are of interest to UK-CAER’s academic, government and industrial research partners. The facility will produce research quantities of FT liquids and finished fuels for subsequent Fuel Quality Testing, Performance and Acceptability. Moreover, the facility is expected to be employed for a range of research and investigations related to: Feed Preparation, Characteristics and Quality; Coal and Biomass Gasification; Gas Clean-up/ Conditioning; Gas Conversion by FT Synthesis; Product Work-up and Refining; Systems Analysis and Integration; and Scale-up and Demonstration. Environmental Considerations - particularly how to manage and reduce carbon dioxide emissions from CBTL facilities and from use of the fuels - will be a primary research objectives. Such a facility has required significant lead time for environmental review, architectural/building construction, and EPC services. UK, with DOE support, has advanced the facility in several important ways. These include: a formal EA/FONSI, and permits and approvals; construction of a building; selection of a range of technologies and vendors; and completion of the upstream process units. The results of this project are the FEED and detailed engineering studies, the alternate configurations and the as-built plant - its equipment and capabilities for future research and demonstration and its adaptability for re-purposing to meet other needs. These are described in some detail in this report, along with lessons learned.

01 COAL, LIGNITE, AND PEAT↗

Autonomous reinforcement learning agents for improving predictions and observations of extreme climate events

Primary Focus Area: This proposal addresses focus area 2, “Predictive modeling through the use of AI techniques.” Science Challenge: Extreme climate events associated with severe weather, coastal and inland flooding, droughts, heat waves and wildfires are expected to increase in frequency and severity in the future. Due to the complexity and chaotic behavior of the climate system, accurately predicting and observing extreme climate events requires a tremendous amount of human intervention to run predictive climate simulations and deploy measurement systems. Extreme events often unfold very quickly, leaving little time to iterate on simulations or re-position instruments. Through reinforcement learning, autonomous AI agents can be designed to make real-time decisions to characterize extreme climate events more efficiently through adaptive models and targeted observations.

54 ENVIRONMENTAL SCIENCES↗

Insightful Workflow For Grid Computing

We developed a workflow adaptation and scheduling system for Grid workflow. The system currently interfaces with and uses the Karajan workflow system. We developed machine learning agents that provide the planner/scheduler with information needed to make decisions about when and how to replan. The Kubrick restructures workflow at runtime, making it unique among workflow scheduling systems. The existing Kubrick system provides a platform on which to integrate additional quality of service constraints and in which to explore the use of an ensemble of scheduling and planning algorithms. This will be the principle thrust of our Phase II work.

97 MATHEMATICS AND COMPUTING↗

Di-CNN: Domain-Knowledge-Informed Convolutional Neural Network for Manufacturing Quality Prediction

In manufacturing, convolutional neural networks (CNNs) are widely used on image sensor data for data-driven process monitoring and quality prediction. However, as purely data-driven models, CNNs do not integrate physical measures or practical considerations into the model structure or training procedure. Consequently, CNNs’ prediction accuracy can be limited, and model outputs may be hard to interpret practically. This study aims to leverage manufacturing domain knowledge to improve the accuracy and interpretability of CNNs in quality prediction. A novel CNN model, named Di-CNN, was developed that learns from both design-stage information (such as working condition and operational mode) and real-time sensor data, and adaptively weighs these data sources during model training. It exploits domain knowledge to guide model training, thus improving prediction accuracy and model interpretability. A case study on resistance spot welding, a popular lightweight metal-joining process for automotive manufacturing, compared the performance of (1) a Di-CNN with adaptive weights (the proposed model), (2) a Di-CNN without adaptive weights, and (3) a conventional CNN. The quality prediction results were measured with the mean squared error (MSE) over sixfold cross-validation. Model (1) achieved a mean MSE of 6.8866 and a median MSE of 6.1916, Model (2) achieved 13.6171 and 13.1343, and Model (3) achieved 27.2935 and 25.6117, demonstrating the superior performance of the proposed model.

47 OTHER INSTRUMENTATION↗

Latent Stochastic Differential Equations for Modeling Quasar Variability and Inferring Black Hole Properties

Quasars are bright and unobscured active galactic nuclei (AGN) thought to be powered by the accretion of matter around supermassive black holes at the centers of galaxies. The temporal variability of a quasar’s brightness contains valuable information about its physical properties. The UV/optical variability is thought to be a stochastic process, often represented as a damped random walk described by a stochastic differential equation (SDE). Upcoming wide-field telescopes such as the Rubin Observatory Legacy Survey of Space and Time (LSST) are expected to observe tens of millions of AGN in multiple filters over a ten year period, so there is a need for efficient and automated modeling techniques that can handle the large volume of data. Latent SDEs are machine learning models well suited for modeling quasar variability, as they can explicitly capture the underlying stochastic dynamics. In this work, we adapt latent SDEs to jointly reconstruct multivariate quasar light curves and infer their physical properties such as the black hole mass, inclination angle, and temperature slope. Our model is trained on realistic simulations of LSST ten year quasar light curves, and we demonstrate its ability to reconstruct quasar light curves even in the presence of long seasonal gaps and irregular sampling across different bands, outperforming a multioutput Gaussian process regression baseline. Our method has the potential to provide a deeper understanding of the physical properties of quasars and is applicable to a wide range of other multivariate time series with missing data and irregular sampling.

79 ASTRONOMY AND ASTROPHYSICS↗

Human factors aspects of control room design

A plan for the design and analysis of a multistation control room is reviewed. It is found that acceptance of the computer based information system by the uses in the control room is mandatory for mission and system success. Criteria to improve computer/user interface include: match of system input/output with user; reliability, compatibility and maintainability; easy to learn and little training needed; self descriptive system; system under user control; transparent language, format and organization; corresponds to user expectations; adaptable to user experience level; fault tolerant; dialog capability user communications needs reflected in flexibility, complexity, power and information load; integrated system; and documentation.

Jenkins, J. P.↗

Lessons learned applying CASE methods/tools to Ada software development projects

This paper describes the lessons learned from introducing CASE methods/tools into organizations and applying them to actual Ada software development projects. This paper will be useful to any organization planning to introduce a software engineering environment (SEE) or evolving an existing one. It contains management level lessons learned, as well as lessons learned in using specific SEE tools/methods. The experiences presented are from Alpha Test projects established under the STARS (Software Technology for Adaptable and Reliable Systems) project. They reflect the front end efforts by those projects to understand the tools/methods, initial experiences in their introduction and use, and later experiences in the use of specific tools/methods and the introduction of new ones.

Blumberg, Maurice H.↗

Lessons from the GP-B Experience for Future Fundamental Physics Missions in Space

Gravity Probe B launched in April 2004 and completed its science data collection in September 2005, with the objective of sub-milliarcsec measurement of two General Relativistic effects on the spin axis orientation of orbiting gyroscopes. Much of the technology required by GP-B has potential application in future missions intended to make precision measurements. The philosophical approach and experiment design principles developed for GP-B are equally adaptable to these mission concepts. This talk will discuss GP-B's experimental approach and the technological and philosophical lessons learned that apply to future experiments in fundamental physics. Measurement of fundamental constants to high precision, probes of short-range forces, searches for equivalence principle violations, and detection of gravitational waves are examples of concepts and missions that will benefit kern GP-B's experience.

Kolodziejczak, Jeffery↗

Predicting Time Series Outputs and Time-to-Failure for an Aircraft Controller Using Bayesian Modeling

Safety of unmanned aerial systems (UAS) is paramount, but the large number of dynamically changing controller parameters makes it hard to determine if the system is currently stable, and the time before loss of control if not. We propose a hierarchical statistical model using Treed Gaussian Processes to predict (i) whether a flight will be stable (success) or become unstable (failure), (ii) the time-to-failure if unstable, and (iii) time series outputs for flight variables. We first classify the current flight input into success or failure types, and then use separate models for each class to predict the time-to-failure and time series outputs. As different inputs may cause failures at different times, we have to model variable length output curves. We use a basis representation for curves and learn the mappings from input to basis coefficients. We demonstrate the effectiveness of our prediction methods on a NASA neuro-adaptive flight control system.

Statistics↗

Towards the Development of a Multi-Agent Cognitive Networking System for the Lunar Environment

This paper details the development of a multi-agent cognitive system intended to optimize networking performance in the lunar environment. NASA’s current concept of the future of lunar communication, LunaNet, outlines a complex network of networks. Challenges such as scalability, interoperability and reliability must first be addressed to successfully fulfill this vision. Machine intelligence can greatly reduce the reliance on human operators and enable efficient operations for tasks such as scheduling and network management. The application of machine learning, artificial intelligence, and other automated decision-making techniques can be used to allow network nodes to intelligently sense and adapt to changes in the environment such as link disruptions, new nodes joining the network, and support for a diverse range of protocols. Cognitive networking seeks to evolve these technologies into an autonomous system with improved science data return, reliability, and scalability. In this paper, we study three main areas a means to further develop cognitive networking capabilities: networking and flight software development, analysis of wireless data for modeling and simulation, and development of algorithms for a multi-agent system.

Rachel Dudukovich↗

PhILMs: Collaboratory on Mathematics and Physics-Informed Learning Machines for Multiscale and Multiphysics Problems

The landscape of computational science and engineering is continually evolving, with the challenge of high-dimensional regression problems standing as a significant hurdle in numerous scientific endeavors. Addressing this challenge, our research, funded by this award, has led to the development of an innovative computational framework known as Probabilistic Partition of Unity Networks (PPOU-Nets). This initiative represents a collaborative effort to harness the potential of mathematics and physics-informed machine learning in tackling multiscale and multiphysics problems prevalent in high-dimensional spaces. Through this work, we have proposed a novel methodology that seamlessly integrates adaptive dimensionality reduction and a mixture of experts model, thereby facilitating a more efficient and accurate approximation of complex functions. This research effort has not only advanced the state of computational science but also opened new avenues for exploration in quantum computing and beyond. This report outlines the motivation, methodology, key findings, and implications of our work, underscoring our contributions to the broader scientific community and the potential pathways for future research.

97 MATHEMATICS AND COMPUTING↗

Dispersion-enhanced sequential batch sampling for adaptive contour estimation

In computer simulation and optimal design, sequential batch sampling offers an appealing way to iteratively stipulate optimal sampling points based upon existing selections and efficiently construct surrogate modeling. Nonetheless, the issue of near duplicates poses tremendous quandary for sequential learning. It refers to the situation that selected critical points cluster together in each sampling batch, which are individually but not collectively informative towards the optimal design. Near duplicates severely diminish the computational efficiency as they barely contribute extra information towards update of the surrogate. To address this issue, we impose a dispersion criterion on concurrent selection of sampling points, which essentially forces a sparse distribution of critical points in each batch, and demonstrate the effectiveness of this approach in adaptive contour estimation. Specifically, we adopt Gaussian process surrogate to emulate the simulator, acquire variance reduction of the critical region from new sampling points as a dispersion criterion, and combine it with the modified expected improvement (EI) function for critical batch selection. The critical region here is the proximity of the contour of interest. This proposed approach is vindicated in numerical examples of a two-dimensional four-branch function, a four-dimensional function with a disjoint contour of interest and a time-delay dynamic system.

97 MATHEMATICS AND COMPUTING↗

Domain-adversarial graph neural networks for Λ hyperon identification with CLAS12

Machine learning methods and in particular Graph Neural Networks (GNNs) have revolutionized many tasks within the high energy physics community. Particularly in the realm of jet tagging, GNNs and domain adaptation have been especially successful. However, applications with lower energy events have not received as much attention. Here, we report on the novel use of GNNs and a domain-adversarial training method to identify Λ hyperon events with the CLAS12 experiment at Jefferson Lab. The GNN method we have developed increases the purity of the Λ yield by a factor of 1.95 and by 1.82 using the domain-adversarial training. This work also provides a good benchmark for developing event tagging machine learning methods for the Λ and other channels at CLAS12 and other experiments, such as the planned Electron Ion Collider.

47 OTHER INSTRUMENTATION↗

Adaptive defuzzification for fuzzy systems modeling

We propose a new parameterized method for the defuzzification process based on the simple M-SLIDE transformation. We develop a computationally efficient algorithm for learning the relevant parameter as well as providing a computationally simple scheme for doing the defuzzification step in the fuzzy logic controllers. The M-SLIDE method results in a particularly simple linear form of the algorithm for learning the parameter which can be used both off- and on-line.

Yager, Ronald R.↗

Efficient Generalized Boundary Detection Using a Sliding Information Distance

In this work, we present a general machine learning algorithm for boundary detection within general signals based on an efficient, accurate, and robust approximation of the universal normalized information distance. Our approach uses an adaptive sliding information distance (SLID) combined with a wavelet-based approach for peak identification to locate the boundaries. Special emphasis is placed on developing an adaptive formulation of SLID to handle general signals with multiple unknown and/or drifting section lengths. Although specialized algorithms may outperform SLID when domain knowledge is available, these algorithms are limited to specific applications and do not generalize. SLID excels in these cases. We demonstrate the versatility and efficacy of SLID on a variety of signal types, including synthetically generated sequences of tokens, binary executables for reverse engineering applications, and time series of seismic events.

42 ENGINEERING↗

Considerations and Lessons Learned from the Use of Gamma Spectroscopy Systems at FUSRAP Project Sites - 17459

This paper will discuss the application of gamma spectroscopy systems for analysis of environmental media samples for parents and daughter progeny of the Natural Uranium and Thorium Decay Series radionuclides. The concentrations and relative ratios of naturally occurring isotopes are used to confirm the presence of FUSRAP-related wastes and for understanding the associated potential dose and health risks. Focus will be given to lessons learned from the set-up and operation of onsite gamma spectroscopy laboratories used to provide stakeholders with defensible and timely results to support FUSRAP site investigations and remediation including adaptive survey design implementation, physical remediation and waste transport, and confirmatory final status survey activities. This paper focuses on the potential impacts to gamma spectroscopy measurement accuracy from the following: - Bias caused by sample matrix effects as well as use of different library reference values between performance evaluation study providers and testing laboratories; - Use of calibration sources with limited energy ranges, and; - Selection of available alternative isotopic decay reference data for the input of line energies and their intensities. These lessons learned include discussing how inter-laboratory measurement differences from blind performance evaluation samples are affected by the manufacturer's measurement method, including equilibrium assumptions; the isotope decay information source library used by the study provider and participating laboratories; and calibration source geometry/density variances from actual sample conditions. Any of these factors can significantly impact the accuracy of reported results data. Deepening the reader's understanding of the various factors that impact the accuracy, interpretation, and overall usability of gamma spectroscopy results data from laboratories will help ensure defensible decision-making and cost-effective mission implementation within FUSRAP and similar remediation programs. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗