Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “generative machine 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 847 records · Page 47

The ab initio amorphous materials database: Empowering machine learning to decode diffusivity

Amorphous materials exhibit unique properties that make them suitable for various applications in science and technology, ranging from optical and electronic devices and solid-state batteries to protective coatings. However, data-driven ex- ploration and design of amorphous materials is hampered by the absence of a com- prehensive database covering a broad chemical space. In this work, we present the largest computed amorphous materials database to date, generated from sys- tematic and accurate ab initio molecular dynamics (AIMD) calculations. We also show how the database can be used in simple machine-learning models to connect properties to composition and structure, here specifically targeting ionic conductiv- ity. These models predict the Li-ion diffusivity with speed and accuracy, offering a cost-effective alternative to expensive density functional theory (DFT) calculations. Furthermore, the process of computational quenching amorphous materials provides a unique sampling of out-of-equilibrium structures, energies, and force landscape, and we anticipate that the corresponding trajectories will inform future work in uni- versal machine learning potentials, impacting design beyond that of non-crystalline materials.

36 MATERIALS SCIENCE↗

Explainable machine learning for incipient anomaly detection in compact molten salt heat exchanger with overlapping feature distributions

High-temperature molten salt-cooled reactors (MSCRs) are a promising next-generation nuclear technology option, offering efficient power conversion and inherent safety features. However, the reliability of these systems depends on the robust operation of heat exchangers (HXs), which are susceptible to failure due to temperature gradients and channel plugging caused by fluid freezing. Conventional monitoring methods, relying on inlet and outlet measurements, lack the spatial resolution needed to detect early-stage faults. We propose a novel design of a compact salt-to-salt matrix-type HX design consisting of interleaved arrays of parallel tubes, with integrated synthetic fiber optic distributed temperature sensing (DTS) to enable localized detection of incipient faults. To evaluate performance of this design, we generate high-fidelity synthetic data using heat transfer computational modeling to simulate channel plugging, and introduce sensor noise for realistic modeling of measurements. The dataset comprises of 97% normal operation and 3% anomaly cases, with each anomaly class representing 1% of the data. These early anomalies result in overlapping temperature profiles between normal and faulty channels, producing a non-separable dataset that challenges traditional classification techniques. We benchmark eight supervised machine learning (ML) models and demonstrate that XGBoost achieves the highest performance. To improve transparency, we develop an explainability framework combining Shapley values and partially ordered sets (POSETs) to quantify and structurally analyze feature importance. This approach identifies both dominant predictors and ambiguous feature relationships, enhancing trust and interpretability. Our results highlight the potential of combining DTS and explainable ML with intelligent feature selection to improve predictive maintenance and ensure operational resilience in advanced nuclear systems.

Prantikos, Konstantinos [Argonne National Laborato↗

Clustering VoIP caller for SPIT identification

Abstract The number of unsolicited and advertisement telephony calls over traditional and Internet telephony has rapidly increased over recent few years. Every year, the telecommunication regulators, law enforcement agencies and telecommunication operators receive a very large number of complaints against these unsolicited, unwanted calls. These unwanted calls not only bring financial loss to the users of the telephony but also annoy them with unwanted ringing alerts. Therefore, it is important for the operators to block telephony spammers at the edge of the network so to gain trust of their customers. In this paper, we propose a novel spam detection system by incorporating different social network features for combating unwanted callers at the edge of the network. To this extent the reputation of each caller is computed by processing call detailed records of user using three social network features that are the frequency of the calls between caller and the callee, the duration between caller and the callee and the number of outgoing partners associated with the caller. Once the reputation of the caller is computed, the caller is then places in a spam and non‐spam clusters using unsupervised machine learning. The performance of the proposed approach is evaluated using a synthetic dataset generated by simulating the social behaviour of the spammers and the non‐spammers. The evaluation results reveal that the proposed approach is highly effective in blocking spammer with 2% false positive rate under a large number of spammers. Moreover, the proposed approach does not require any change in the underlying VoIP network architecture, and also does not introduce any additional signalling delay in a call set‐up phase. Copyright © 2016 John Wiley & Sons, Ltd.

Azad, Muhammad Ajmal↗

An Algorithm to Generate Synthetic 3D Microstructures from 2D Exemplars

The inverse problem of constructing 3D microstructures from 2D data is an area of active research within the materials science community. This paper presents the implementation of a robust, computationally efficient algorithm, Hierarchical Algorithm for the Reconstruction of Exemplars (HARE), written in Python to reconstruct 3D features in a given microstructure from up to three orthogonal 2D exemplars using nearest neighbor matching to reproduce feature qualities, such as shape, size, and distribution. HARE’s feature sampling implements histogram reweighting to avoid both over- and undersampling. A neighborhood voting scheme allows each pixel to provisionally affect its neighbors according to its weight. Finally, the algorithm is presently configured for two-phase materials and is being extended to accommodate multiple phases. HARE is a convenient and robust base from which to generate statistically representative synthetic microstructures for use in multi-scale modeling or machine learning applications to support advanced manufacturing and materials discovery.

36 MATERIALS SCIENCE↗

CASM — A software package for first-principles based study of multicomponent crystalline solids

CASM is a software package that enables first-principles based studies of crystalline materials. It has been designed to treat coupled chemical, mechanical, vibrational, and magnetic degrees of freedom to determine ground state and finite temperature properties of crystals. The symmetry of the underlying parent crystal structure is used to enumerate perturbations of the parent crystal structure and generate derivative structures which can be input to first-principles calculations in order to explore the ground state energy landscape. CASM algorithmically constructs cluster expansions that fully couple discrete and continuous degrees of freedom, and generates highly efficient code to evaluate the cluster expansion basis functions. Widely used machine learning methods are integrated for fitting expansion coefficients to first-principle calculations. The fully parameterized cluster expansions can be combined with (kinetic) Monte Carlo methods to calculate finite temperature thermodynamic and kinetic properties. CASM Alloy Manager identifies distinct parent crystal structures in an alloy system, creating individual projects for each, and integrating the results. The integrated infrastructure facilitates the linkage between first-principles statistical mechanics predictions with higher length scale computational methods, such as phase field simulations. In conclusion, CASM projects are designed to be easy to share in repositories, to re-use, and to extend to include additional chemical species or types of degrees of freedom.

36 MATERIALS SCIENCE↗

Detecting control system misbehavior by fingerprinting programmable logic controller functionality

In recent years, attacks such as the Stuxnet malware have demonstrated that cyberattacks against control systems cause extensive damage. These attacks can result in physical damage to the networked systems under their control. In this paper, we discuss our approach for detecting such attacks by distinguishing between programs running on a programmable logic controller (PLC) without having to monitor communications. Using power signatures generated by an attached, high-frequency power measurement device, we can identify what a PLC is doing and when an attack may have altered what the PLC should be doing. To accomplish this, we generated labeled data for testing our methods and applied feature engineering techniques and machine learning models. The results of this research demonstrate that Random Forests and Convolutional Neural Networks classify programs with up to 98% accuracy for major program differences and 84% accuracy for minor differences. Our results can be used for both online and offline applications.

97 MATHEMATICS AND COMPUTING↗

Improving Low‐Cloud Fraction Prediction Through Machine Learning

Abstract In this study, we evaluated the performance of machine learning (ML) models (XGBoost) in predicting low‐cloud fraction (LCF), compared to two generations of the community atmospheric model (CAM5 and CAM6) and ERA5 reanalysis data, each having a different cloud scheme. ML models show a substantial enhancement in predicting LCF regarding root mean squared errors and correlation coefficients. The good performance is consistent across the full spectrums of atmospheric stability and large‐scale vertical velocity. Employing an explainable ML approach, we revealed the importance of including the amount of available moisture in ML models for representing spatiotemporal variations in LCF in the midlatitudes. Also, ML models demonstrated marked improvement in capturing the LCF variations during the stratocumulus‐to‐cumulus transition (SCT). This study suggests ML models' great potential to address the longstanding issues of “too few” low clouds and “too rapid” SCT in global climate models.

Geology↗

Grooved electrodes for high-power-density fuel cells

Proton exchange membrane fuel cells (PEMFCs) are leading candidates to decarbonize the transport sector, but widespread deployment will require improvements in lifetime, fuel economy and cost. Here we present the grooved electrode, an alternative electrode structure that enhances PEMFC performance and durability by coupling high ionomer (ion-conducting binder) content for improved H + transport with grooves for rapid O 2 transport. Grooved electrodes provide up to 50% higher performance than state-of-the-art conventional electrodes under standard operating conditions. Fuel cell diagnostics combined with multiphysics modelling demonstrate that grooved electrodes provide facile O 2 transport despite their high ionomer content, enabling improved reaction rate uniformity. Grooved electrodes also provide improved durability, with less performance loss after carbon corrosion compared with baseline electrodes. Machine learning analysis demonstrates the potential to further optimize grooved structures for next-generation PEMFCs with enhanced performance and durability, enabling smaller and cheaper fuel cell stacks with higher fuel efficiency.

25 ENERGY STORAGE↗

The case for digital twins in metal additive manufacturing

The digital twin (DT) is a relatively new concept that is finding increased acceptance in industry. A DT is generally considered as comprising a physical entity, its virtual replica, and two-way digital data communications in-between. Its primary purpose is to leverage the process intelligence captured within digital models—or usually their faster-solving surrogates—towards generating increased value from the physical entities. The surrogate models are created using machine learning based on data obtained from the field, experiments and digital models, which may be physics-based or statistics-based. Anomaly detection and correction, and diagnostic closed-loop process control are examples of how a process DT can be deployed. In the manufacturing industry, its use can achieve improvements in product quality and process productivity. Metal additive manufacturing (AM) stands to gain tremendously from the use of DTs. This is because the AM process is inherently chaotic, resulting in poor repeatability. However, a DT acting in a supervisory role can inject certainty into the process by actively keeping it within bounds through real-time control commands. Closed-loop feedforward control is achieved by observing the process through sensors that monitor critical parameters and, if there are any deviations from their respective optimal ranges, suitable corrective actions are triggered. The type of corrective action (e.g. a change in laser power or a modification to the scanning speed) and its magnitude are determined by interrogating the surrogate models. Because of their artificial intelligence (AI)-endowed predictive capabilities, which allow them to foresee a future state of the physical twin (e.g. the AM process), DTs proactively take context-sensitive preventative steps, whereas traditional closed-loop feedback control is usually reactive. Apart from assisting a build process in real-time, a DT can help with planning the build of a part by pinpointing the optimum processing window relevant to the desired outcome. Again, the surrogate models are consulted to obtain the required information. In this article, we explain how the application of DTs to the metal AM process can significantly widen its application space by making the process more repeatable (through quality assurance) and cheaper (by getting builds right the first time).

36 MATERIALS SCIENCE↗

The seventh blind test of crystal structure prediction: structure ranking methods

A seventh blind test of crystal structure prediction has been organized by the Cambridge Crystallographic Data Centre. The results are presented in two parts, with this second part focusing on methods for ranking crystal structures in order of stability. The exercise involved standardized sets of structures seeded from a range of structure generation methods. Participants from 22 groups applied several periodic DFT-D methods, machine learned potentials, force fields derived from empirical data or quantum chemical calculations, and various combinations of the above. In addition, one non-energy-based scoring function was used. Results showed that periodic DFT-D methods overall agreed with experimental data within expected error margins, while one machine learned model, applying system-specific AIMnet potentials, agreed with experiment in many cases demonstrating promise as an efficient alternative to DFT-based methods. For target XXXII, a consensus was reached across periodic DFT methods, with consistently high predicted energies of experimental forms relative to the global minimum (above 4 kJ mol −1 at both low and ambient temperatures) suggesting a more stable polymorph is likely not yet observed. The calculation of free energies at ambient temperatures offered improvement of predictions only in some cases (for targets XXVII and XXXI). Several avenues for future research have been suggested, highlighting the need for greater efficiency considering the vast amounts of resources utilized in many cases.

Chemistry↗

A Data-Driven Framework for Residential Electric Vehicle Charging Load Profile Generation

Residential electric vehicle charging load profile is indispensable to achieve reliable control strategies for mitigating negative effects on power distribution system due to emerging electrified transportation. This paper introduces a data-driven framework of charging load profile generation for residential plug-in electric vehicles. Real world historical residential charging behavior data is utilized to construct empirical charging decision making model by using machine learning algorithm. A multiple channels method with kernel density estimation is proposed to construct probability density functions for estimating charging duration based on parking duration. A generation algorithm considering parking time and travel demand dependency is introduced to generate residential charging behaviors. This framework is extensible to generate various charging load profiles and simulate varied residential charging scenarios under different number of households and charging rates. This will be crucial for designing and validating residential charging control strategies.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Training data selection for event classification in a highly variable environment

A problem of interest for nuclear nonproliferation is monitoring activities at nuclear facilities, where proliferation events may only take place a few times and often under variable conditions. Machine learning has revolutionized data analytics by enabling the use of measurable signatures to generate predictive models of facility operations. However, traditional methods for training these models require large, reliable data sets with labeled observations, a challenge for nonproliferation. Highly variable conditions further complicate this as events from training data may have occurred in conditions quite different from the event of interest. Our hypothesis is that when events occur in a highly variable environment, careful training data selection for each test event could outperform the standard approach of using all available training data. We developed a method to optimize training data selection for the given test event and applied it to predicting the power level of the High Flux Isotope Reactor (HFIR) at Oak Ridge National Laboratory. In this study, the reactor startup exhibits variability between occurrences due to natural variability in environmental conditions and operational procedures. Using a combination of analysis techniques, a similitude assessment was performed on data collected from HFIR to isolate clusters that were optimal for training a predictive model. Concepts such as dynamic time warping and Jaccard similarity were used in conjunction with clustering analysis. In order to validate this approach, the model was trained on every combination of unique training events and the predictive performance was compared to the performance using a subset of the training data selected by isolated clusters found through the similitude assessment.

Iyer, A↗

Interactive Machine Learning in Data Exploitation

The goal of interactive machine learning is to help scientists and engineers exploit more of their specialized data in less time. Interactive machine learning focuses on methods that empower domain experts to control and direct machine learning tools from within the deployed environment, whereas traditional machine learning does this in the development environment. This difference allows interactive machine learning systems to be more responsive, more accurate and cheaper to develop and maintain. This article provides a basic introduction to the main components and tries to untangle the many ideas that must be combined to produce practical interactive learning systems. It also describes recent developments in machine learning that have significantly advanced the theoretical and practical foundations for the next generation of interactive tools.

Information Science↗

AI Denoising to Accelerate Detector Simulation

Detector simulation is critical to experimental HEP; however this simulation (commonly done through toolkits such as Geant4) is computationally intensive. Performance can be improved somewhat through technical optimization, but more is needed. Using machine learning (ML) to accelerate simulation is a promising field, however efforts to use generative adversarial networks (GANs) or optimized autoencoders have faced issues. Using convolutional neural networks (CNNs) for denoising has been successful in non-HEP applications such as image processing. This poster investigates the efficacy of using CNNs to denoise Geant4 simulations. This could increase the accuracy of simulations performed under settings designed to increase computational efficiency.

Franklin, Lena↗

Enabling the Next Generation of Smart Sensors in Coal Fired Power Plants using Cellular 5G Technology

An important need for coal fired power plants is the ability to monitor multiple systems with ease and accuracy. Common implementations of these monitoring systems come with drawbacks due to the nature of coal fired power plants. Harsh environments, High Temperatures, and lots of RF (Radio Frequency) noise can create issues for accurately recording and transmitting data across wireless signals. In addition, as renewable energy sources come online, existing fossil fueled plants will need to operate more flexibly with their maintenance schedules outside of standard conditions. Therefore, additional sensing and control mechanisms need placed in existing plants to provide operators with more information such that maintenance decisions can be made well in advance of failures. A solution to this problem is the Next Generation of Smart Sensors, which leverages the power of 5G cellular signals and machine learning to overcome the myriad of problems with current implementations

20 FOSSIL-FUELED POWER PLANTS↗

Dataset, Code, and Models for Training Deep Learning Potentials for Low Temperature Plasma-Surface Interactions

This repository contains datasets, training scripts, and finished models, and test simulations used in the development of DeepREBO— a machine-learned interatomic potential trained to emulate the REBO2 empirical potential. The data was generated to study deep potential development for simulations of plasma-surface interactions. It uses an active learning framework, starting from a minimal dataset and iteratively expanding it. Included are those generated datasets, the trained models, and simulations used to evaluate the performance of the training process. This resource supports reproducibility and provides a reference framework for training deep potentials in plasma-surface interaction studies.

active learning↗

Gaussian Multiple Instance Learning Approach for Mapping the Slums of the World Using Very High Resolution Imagery

In this paper, we present a computationally efficient algo- rithm based on multiple instance learning for mapping infor- mal settlements (slums) using very high-resolution remote sensing imagery. From remote sensing perspective, infor- mal settlements share unique spatial characteristics that dis- tinguish them from other urban structures like industrial, commercial, and formal residential settlements. However, regular pattern recognition and machine learning methods, which are predominantly single-instance or per-pixel classi- fiers, often fail to accurately map the informal settlements as they do not capture the complex spatial patterns. To overcome these limitations we employed a multiple instance based machine learning approach, where groups of contigu- ous pixels (image patches) are modeled as generated by a Gaussian distribution. We have conducted several experi- ments on very high-resolution satellite imagery, represent- ing four unique geographic regions across the world. Our method showed consistent improvement in accurately iden- tifying informal settlements.

Vatsavai, Raju↗

A Measurement of the Pion-Energy Dependence of Muon Neutrino Charged-Current Scattering to Final States With One Charged Pion in NOvA

The study of neutrino oscillations is a main priority for particle physics as the most immediately tractable lever on physics beyond the Standard Model. In particular, more insight into violation of the combined symmetry of charge-conjugation plus parity could yield clues to the origin of matter-antimatter asymmetry, and in some theoretical frameworks the neutrino mass could give insights into dark matter. NOvA is a long-baseline accelerator neutrino experiment with both a near and far detector that seeks to measure several of the parameters of the neutrino mixing matrix, as well as carry out a broad program of additional physics. NOvA has innovated and developed a variety of techniques in the space of neutrino physics, including expanding the use of machine learning techniques in reconstruction. NOvA has also set the stage for the next generation US-based long baseline experiment, DUNE. This dissertation details the creation of a new neutrino interaction vertex reconstruction package for NOvA, which offers enormous improvements in accuracy above the previous vertexer it replaces. This is accomplished using a Convolutional Visual Network trained on large datasets of simulated events. The vertexer is validated thoroughly against data. This dissertation also presents a cross-section measurement for $\nu_\mu + N \rightarrow \mu^- + 1\pi^\pm + X$ (where X does not include additional charged pions) binned in pion kinetic energy, a challenging measurement in general and for NOvA in particular. Measurements like this provide crucial inputs for neutrino interaction models utilized by neutrino oscillation experiments. This measurement in particular is important to the energy ranges relevant for DUNE in a band not well-covered by any other experiment.

Ewart, Erin [Indiana U., Bloomington (main)]↗