Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “machine”

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 883 records · Page 49

Calculation of machine precision second order derivatives using dual-complex numbers

It is well known that both complex and dual numbers can be employed to obtain machine precision first-order derivatives; however, neither, on their own, can compute machine precision 2nd order derivatives. To address this limitation, it is demonstrated in this paper that combined dual-complex numbers can be used to compute machine precision 1st and 2nd order derivatives. The dual-complex approach is simpler than utilizing multicomplex or hyper-dual numbers as existing dual libraries can be used as is or easily augmented to accept complex numbers, and the complexity of developing, integrating, and deploying multicomplex or hyper-dual libraries is avoided. The efficacy of this approach is demonstrated for both univariate and multivariate functions. Finally, source code examples using the Python, Julia, and Mathematica languages are provided as supplemental material.

97 MATHEMATICS AND COMPUTING↗

Graph coarsening: from scientific computing to machine learning

Abstract The general method of graph coarsening or graph reduction has been a remarkably useful and ubiquitous tool in scientific computing and it is now just starting to have a similar impact in machine learning. The goal of this paper is to take a broad look into coarsening techniques that have been successfully deployed in scientific computing and see how similar principles are finding their way in more recent applications related to machine learning. In scientific computing, coarsening plays a central role in algebraic multigrid methods as well as the related class of multilevel incomplete LU factorizations. In machine learning, graph coarsening goes under various names, e.g., graph downsampling or graph reduction. Its goal in most cases is to replace some original graph by one which has fewer nodes, but whose structure and characteristics are similar to those of the original graph. As will be seen, a common strategy in these methods is to rely on spectral properties to define the coarse graph.

Chen, Jie↗

Macroscopic Traffic Modeling Using Probe Vehicle Data: A Machine Learning Approach

Abstract The macroscopic fundamental diagram (MFD) captures an orderly relationship among traffic flow, density, and speed at the network level. It is a simple yet powerful tool for modeling traffic dynamics in large urban networks with broad application in traffic control and management. However, empirically derived MFDs in urban regions require high-resolution traffic data from the network. Having the network flow and vehicular density estimated at the (granular) census tract level using vehicle probe data, we apply machine learning methods to predict the MFDs across U.S. urban areas and capture the impacts of location-specific input features on the network flow–density relationships at a large scale. The results show that, among the four tested machine learning approaches (Random Forest, XGBoost, Support Vector Machine, and Neural Network), XGBoost delivers the best performance in predicting network traffic flow based on vehicular density and location attributes. Using interaction Shapley Additive explanation (SHAP) values and partial correlation analysis, we examine the factors influencing MFD shapes across different locations. Our empirical findings reveal that across U.S. urban areas, network topology, transportation infrastructure, and land use are primary factors shaping MFD curves, while demand and trip-related factors play a lesser role. Specifically, higher ranking roads, centrality, and development levels correlate positively with network capacity and critical density, whereas negative associations are observed for network connectivity, mixed-use development, and road roughness levels.

Jin, Ling↗

Quantum circuit fidelity estimation using machine learning

The computational power of real-world quantum computers is limited by errors. When using quantum computers to perform algorithms which cannot be efficiently simulated classically, it is important to quantify the accuracy with which the computation has been performed. In this work, we introduce a machine learning-based technique to estimate the fidelity between the state produced by a noisy quantum circuit and the target state corresponding to ideal noise-free computation. Our machine learning model is trained in a supervised manner, using smaller or simpler circuits for which the fidelity can be estimated using other techniques like direct fidelity estimation and quantum state tomography. Here we demonstrate that, for simulated random quantum circuits with a realistic noise model, the trained model can predict the fidelities of more complicated circuits for which such methods are infeasible. In particular, we show that the trained model may make predictions for circuits with higher degrees of entanglement than were available in the training set and that the model may make predictions for non-Clifford circuits even when the training set included only Clifford-reducible circuits. This empirical demonstration suggests classical machine learning may be useful for making predictions about beyond-classical quantum circuits for some non-trivial problems.

97 MATHEMATICS AND COMPUTING↗

Classical information driven quantum dot thermal machines

We analyze the transient response of quantum dot thermal machines that can be driven by hyperfine interaction acting as a source of classical information. Our setup comprises a quantum dot coupled to two contacts that drive heat flow while coupled to a nuclear spin bath. The quantum dot thermal machines operate both as batteries and as engines, depending on the parameter range. The electrons in the quantum dot interact with the nuclear spins via hyperfine spin-flip processes as typically seen in solid state systems such as GaAs quantum dots. The hyperfine interaction in such systems, which is often treated as a deterrent for quantum information processing, can favorably be regarded as a driving agent for classical information flow into a heat engine setup. We relate this information flow to Landauer’s erasure of the nuclear spin bath, leading to a battery operation. We further demonstrate that the setup can perform as a transient power source even under a voltage bias across the dot. Focusing on the transient thermoelectric operation, our analysis clearly indicates the role of Landauer’s erasure to deliver a higher output power than a conventional quantum dot thermoelectric setup and an efficiency greater than that of an identical Carnot cycle in steady state, which is consistent with recently proposed bounds on efficiency for systems subject to a feedback controller. The role of nuclear spin relaxation processes on these aspects is also studied. Finally, we introduce the Coulomb interaction in the dot and analyze the transient thermoelectric response of the system. Our results elaborate on the effective use of somewhat undesirable scattering processes as a non-equilibrium source of Shannon information flow in thermal machines and the possibilities that may arise from the use of a quantum information source.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Time-series machine-learning error models for approximate solutions to parameterized dynamical systems

This work proposes a machine-learning framework for modeling the error incurred by approximate solutions to parameterized dynamical systems. In particular, we extend the machine-learning error models (MLEM) framework proposed in Ref. Freno and Carlberg (2019) to dynamical systems. The proposed Time-Series Machine-Learning Error Modeling (T-MLEM) method constructs a regression model that maps features – which comprise error indicators that are derived from standard a posteriori error-quantification techniques – to a random variable for the approximate-solution error at each time instance. The proposed framework considers a wide range of candidate features, regression methods, and additive noise models. We consider primarily recursive regression techniques developed for time-series modeling, including both classical time-series models (e.g., autoregressive models) and recurrent neural networks (RNNs), but also analyze standard non-recursive regression techniques (e.g., feed-forward neural networks) for comparative purposes. Finally, numerical experiments conducted on multiple benchmark problems illustrate that the long short-term memory (LSTM) neural network, which is a type of RNN, outperforms other methods and yields substantial improvements in error predictions over traditional approaches.

42 ENGINEERING↗

Influence of surface modifications on pitting corrosion behavior of nickel-base alloy 718. Part 1: Effect of machine hammer peening

The effect of surface modifications induced by machine hammer peening on pitting corrosion behavior of nickel-base alloy 718 in a 3.5 wt.% NaCl solution is investigated. Severe work hardening and high compressive residual stress are generated with surface smoothing and microstructure evolution in terms of formation of nano-grains and nano-twins in the near surface region after machine hammer peening. Electrochemical tests results show that machine hammer peening has a beneficial influence on the corrosion resistance, indicated by a significant increase of the critical pitting potential (+134 mV) accompanied with lower corrosion current density and higher polarization resistance.

36 MATERIALS SCIENCE↗

Predicting Future Hourly Residential Electrical Consumption: A Machine Learning Case Study

Whole building input models for energy simulation programs are frequently created in order to evaluate specific energy savings potentials. They are also often utilized to maximize cost-effective retrofits for existing buildings as well as to estimate the impact of policy changes toward meeting energy savings goals. Traditional energy modeling suffers from several factors, including the large number of inputs required to characterize the building, the specificity required to accurately model building materials and components, simplifying assumptions made by underlying simulation algorithms, and the gap between the as-designed and as-built building. Prior works have attempted to mitigate these concerns by using sensor-based machine learning approaches to model energy consumption. However, a majority of these prior works focus only on commercial buildings. The works that focus on modeling residential buildings primarily predict monthly electrical consumption, while commercial models predict hourly consumption. This means there is not a clear indicator of which techniques best model residential consumption, since these methods are only evaluated using low-resolution data. We address this issue by testing seven different machine learning algorithms on a unique residential data set, which contains 140 different sensors measurements, collected every 15 minutes. In addition, we validate each learner's correctness on the ASHRAE Great Energy Prediction Shootout, using the original competition metrics. Our validation results confirm existing conclusions that Neural Network-based methods perform best on commercial buildings. However, the results from testing our residential data set show that Feed Forward Neural Networks, Support Vector Regression (SVR), and Linear Regression methods perform poorly, and that Hierarchical Mixture of Experts (HME) with Least Squares Support Vector Machines (LS-SVM) performs best - a technique not previously applied to this domain.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Use of system code to estimate equilibrium tritium inventory in fusion DT machines, such as ARIES-AT and components testing facilities

ITER is under construction and will begin operation in 2020. This is the first 500 MWfusion class DT device, and since it is not going to breed tritium, it will consume most of the limited supply of tritium resources in the world. Yet, in parallel, DT fusion nuclear component testing machines will be needed to provide technical data for the design of DEMO. It becomes necessary to estimate the tritium burn-up fraction and corresponding initial tritium inventory and the doubling time of these machines for the planning of future supply and utilization of tritium. With the use of a system code, tritium burn-up fraction and initial tritium inventory for steady state DT machines can be estimated. Estimated tritium burn-up fractions of FNSF-AT, CFETR-R and ARIES-AT are in the range of 1–2.8%. Corresponding total equilibrium tritium inventories of the plasma flow and tritium processing system, and with the DCLL blanket option are 7.6 kg, 6.1 kg, and 5.2 kg for ARIES-AT, CFETR-R and FNSF-AT, respectively.

99 GENERAL AND MISCELLANEOUS↗

Normal or abnormal? Machine learning for the leakage detection in carbon sequestration projects using pressure field data

The international commitments for atmospheric carbon reduction will require a rapid increase in carbon capture and storage (CCS) projects. The key to any successful CCS project lies in the long term storage and prevention of leakage of stored carbon dioxide (CO 2 ). In addition to being a greenhouse gas, CO 2 leaks reaching the surface can accumulate in low-lying areas resulting in a serious health risk. Among several alternatives, some of the more promising CCS storage formations are depleted oil and gas reservoirs, where the reservoirs had good geological seals prior to hydrocarbon extraction. With more CCS wells coming online, it is imperative to implement permanent, automated monitoring tools. We apply machine learning models to automate the leakage detection process in carbon storage reservoirs using rates of (CO 2 ) injection and pressure data measured by simple harmonic pulse testing (HPT). To validate the feasibility of this machine learning based workflow, we use data from HPT experiments carried out in the Cranfield oil field, Mississippi, USA. The data consist of a series of pulse tests conducted with baseline parameters and with an artificially introduced leak. Here, in this study, we pose the leakage detection task as an anomaly detection problem where deviation from the predicted behavior indicates leaks in the reservoir. Results show that different machine learning architectures such as multi-layer feed forward network, Long Short-Term Memory, and convolutional neural network are able to identify leakages and can provide early warning. These warnings can then be used to take remedial measures.

58 GEOSCIENCES↗

Machine Learning Models for Binary Molecular Classification using VUV Absorption Spectra

Machine learning methods were combined with differential absorption spectroscopy measurements in the vacuum-ultraviolet region (5.167 – 9.920 eV) in order to develop predictive capabilities for inferring molecular structure from the spectra. Several types of species were analyzed and, for modeling purposes, were defined using a single classification: (1) alkane, (2) conjugation with oxygen (e.g. diacetyl, ethyl vinyl ether), (3) non-conjugated alkene (e.g. 1-butene, 1,4-cyclohexadiene), (4) oxygen-containing (e.g. 1-butanol, tetrahydrofuran), or (5) cyclic (e.g. cyclopentane, cyclohexanone). The latter molecular classification excluded cyclic ethers. Several modeling methods were employed in the analysis of 102 absorption spectra, 24 of which were measured for the first time. The primary objective was to identify suitable methods that enable accurate predictions of molecular structure classifications with minimized statistical uncertainties. Rather than identifying a single, unifying method to reliably predict molecular structure contributions to VUV absorption spectra, coordination is required among a particular method, the type of molecular structure detail (e.g. conjugation), and absorption region of interest. The latter is accomplished using a binning approach, wherein absorption regions of ~0.5 eV were utilized rather than the entire ~4.8 eV range. Photon energy binning enabled analysis of region-specific predictions of accuracy, precision, and recall. The outcome from the binning approach is that, rather than utilizing the entire spectrum, optimal determination of molecular structure using machine learning methods depends on the absorption region. Furthermore, the present work provides separate machine learning models for each molecular classification, which enables the identification of multi-functional species relevant to atmospheric chemistry and combustion chemistry, where isomer-resolved speciation is critical to understanding complex reaction networks.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Enhancing nanoscale SEM image segmentation and reconstruction with crystallographic orientation data and machine learning

Current methods of image segmentation and reconstructions from scanning electron micrographs can be inadequate for resolving nanoscale gaps in composite materials (1–20 nm). Such information is critical to both accurate material characterizations and models of piezoresistive response. The current work proposes the use of crystallographic orientation data and machine learning for enhancing this process. It is first shown how a machine learning algorithm can be used to predict the connectivity of nanoscale grains in a Nickel nanostrand/epoxy composite. This results in 71.9% accuracy for a 2D algorithm and 62.4% accuracy in 3D. Finally, it is demonstrated how these algorithms can be used to predict the location of gaps between distinct nanostrands — gaps which would otherwise not be detected with the sole use of a scanning electron microscope. - Highlights: • A method is proposed for enhancing the segmentation/reconstruction of SEM images. • 3D crystallographic orientation data from a nickel nanocomposite is collected. • A machine learning algorithm is used to detect trends in adjacent grains. • This algorithm is then applied to predict likely regions of nanoscale gaps. • These gaps would otherwise be unresolved with the sole use of an SEM.

36 MATERIALS SCIENCE↗

Effect of starting microstructure upon the nucleation sites and distribution of graphite particles during a graphitising anneal of an experimental medium-carbon machining steel

The potential for using graphite particles as an internal lubricant during machining is considered. Graphite particles were found to form during graphitisation of experimental medium-carbon steel alloyed with Si and Al. The graphite nucleation sites were strongly influenced by the starting microstructure, whether ferrite–pearlite, bainite or martensite, as revealed by light and electron microscopy. Favourable nucleation sites in the ferrite–pearlite starting microstructure were, not unexpectedly, found to be located within pearlite colonies, no doubt due to the presence of abundant cementite as a source of carbon. In consequence, the final distribution of graphite nodules in ferrite–pearlite microstructures was less uniform than for the bainite microstructure studied. In the case of martensite, this study found a predominance of nucleation at grain boundaries, again leading to less uniform graphite dispersions. - Highlights: • Metallography of formation of graphite particles in experimental carbon steel. • Potential for using graphite in steel as an internal lubricant during machining. • Microstructure features expected to influence improved machinability studied. • Influence of pre-anneal starting microstructure on graphite nucleation sites. • Influence of pre-anneal starting microstructure on graphite distribution. • Potential benefit is new free-cutting steel compositions without e.g. Pb alloying.

36 MATERIALS SCIENCE↗

Superplasticity and machinability in a four-phase ceramic

Highlights: ► New fine grained ceramic composites of ZrO{sub 2}–Al{sub 2}O{sub 3}–Mg{sub 2}AlO{sub 4}–LaPO{sub 4} were synthesized. ► At high temperatures, deformation strain rates are comparable to superplastic TZP. ► These are machinable at room temperature using conventional tungsten drill bits. ► Magnetoplumbite can form under certain process conditions. -- Abstract: Fine grained four-phase ceramic materials were fabricated to have a combination of high temperature superplasticity and room temperature machinability. The composite ceramics were made of 25 vol.% 3Y-TZP/8YSZ–25 vol.% Al{sub 2}O{sub 3}–25 vol.% MgAl{sub 2}O{sub 4}–25 vol.% LaPO{sub 4}, using LaPO{sub 4} powders fabricated in-house. X-ray diffraction and scanning electron microscopy revealed that the grain size for the best mixed samples was in the range of 600 nm, tetragonal ZrO{sub 2} transformed into monoclinic, and reactions in the sintered samples produced a new phase, magnetoplumbite (LaMgAl{sub 11}O{sub 19}) with lath-like grains. The formation of magnetoplumbite was facilitated by the presence of yttria and by a liquid phase generated at elevated temperatures. These four-phase ceramics had a maximum hardness of 12 GPa and fracture toughness was no more than 3 MPa√m. Deformation rates at 1400 °C under 40 MPa stress were in the superplastic range of 10{sup −3} s{sup −1} for most compositions. These four-phase ceramics were machinable as demonstrated using conventional tungsten drill bits.

36 MATERIALS SCIENCE↗

Nano-machining, surface analysis and emittance measurements of a copper photocathode at SPARC_LAB

R&D activity on Cu photocathodes is under development at the SPARC_LAB test facility to fully characterize each stage of the photocathode “life” and to have a complete overview of the photoemission properties in high brightness photo-injectors. The nano(n)-machining process presented here consists in diamond milling, and blowing with dry nitrogen. This procedure reduces the roughness of the cathode surface and prevents surface contamination introduced by other techniques, such as polishing with diamond paste or the machining with oil. Both high roughness and surface contamination cause an increase of intrinsic emittance and consequently a reduction of the overall electron beam brightness. To quantify these effects, we have characterized the photocathode surface in terms of roughness measurement, and morphology and chemical composition analysis by means of Scanning Electron Microscopy (SEM), Energy Dispersive Spectroscopy (EDS), and Atomic Force Microscopy (AFM) techniques. The effects of n-machining on the electron beam quality have been also investigated through emittance measurements before and after the surface processing technique. Finally, we present preliminary emittance studies of yttrium thin film on Cu photocathodes.

36 MATERIALS SCIENCE↗

Machine learning framework for predicting uranium enrichments from M400 CZT gamma spectra

A machine learning framework was developed for predicting uranium enrichments from M400 CZT gamma spectra. This framework leverages the availability of a large amount of measured M400 gamma spectra and uses a recently updated version of Gamma Detector Response and Analysis Software (GADRAS) for gamma spectrum analysis and generation. It also leverages the existing machine learning modules in Python for gamma spectrum data processing, curation, model training, benchmarking, and optimization of the deep machine learning models. The framework is used to develop a deep learning model to analyze gamma spectra from a set of U 3 O 8 samples with enrichments ranging from 0.31 to 93.17% and UF 6 cylinders with enrichments ranging from 0.2 to 4.95%, and the model performance is tested using a set of measured spectra and the respective declared enrichment values. Results show that the model can correctly classify 99.35% of the U 3 O 8 sample enrichments, and can predict the samples’ enrichments within an average absolute error of 0.099% (in percentage points of enrichment). For the UF 6 cylinders, the average absolute error was approximately 0.03%, with an accuracy of 98% in classifying discrete enrichment values of UF 6 samples. Finally, the results also show that the model has performed significantly better in terms of predicting enrichments in UF 6 cylinders based on measured gamma spectra than the GEM code, with a standard deviation (of the relative errors) of 2.23% (compared with the 11.51% value for the GEM code) based on results from a set of test data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Machine vision based particle size and size distribution determination of airborne dust particles of wood and bark pellets

Dust management strategies in industrial environment, especially of airborne dust, require quantification and measurement of size and size distribution of the particles. Advanced specialized instruments that measure airborne particle size and size distribution apply indirect methods that involve light scattering, acoustic spectroscopy, and laser diffraction. In this research, we propose a simple and direct method of airborne dust particle dimensional measurement and size distribution analysis using machine vision. The method involves development of a user-coded ImageJ plugin that measures particle length and width and analyzes size distribution of particles based on particle length from high-resolution scan images. Test materials were airborne dust from soft pine wood sawdust pellets and ground pine tree bark pellets. Subsamples prepared by dividing the actual dust using 230 mesh (63 m) sieve were analyzed as well. A flatbed document scanner acquired the digital images of the dust particles. Proper sampling, layout of dust particles in singulated arrangement, good contrast smooth background, high resolution images, and accurate algorithm are essential for reliable analysis. A halo effect around grey-scale images ensured correct threshold limits. The measurement algorithm used Feret s diameter for particle length and pixel-march technique for particle width. Particle size distribution was analyzed in a sieveless manner after grouping particles according to their distinct lengths, and several significant dimensions and parameters of particle size distribution were evaluated. Results of the measurement and analysis were presented in textual and graphical formats. The developed plugin was evaluated to have a dimension measurement accuracy in excess of 98.9% and a computer speed of analysis of <8 s/image. Arithmetic mean length of actual wood and bark pellets airborne dust particles were 0.1138 0.0123 and 0.1181 0.0149 mm, respectively. The airborne dust particles of wood and bark pellets can be described as non uniform, finer particles dominated, very finely skewed with positive skewness, leptokurtic, and very well sorted category. Experimental mechanical sieving and machine vision methods produced comparable particle size distribution. The limitations and merits of using the machine vision technique for the measurement of size and size distribution of fine particles such as airborne dust were discussed.

09 BIOMASS FUELS↗

Toward Machine Learning-Driven Mass Spectrometric Identification of Trichothecenes in the Absence of Standard Reference Materials

While a significant body of work exists on the detection of commonly known trichothecene toxins, biological, environmental, and other transformational processes can generate many under-characterized and unknown modified trichothecenes. Lacking both analytical reference standards and associated mass spectral databases, identification of these modified compounds reflects both a challenge and a critical gap from forensic and public health perspectives. Here we report here the application of machine learning (ML) techniques toward identification of discriminative fragment ions from mass spectrometric data that can be exploited to detect evidence of type A and B trichothecenes. The goal of this work is to establish a new method for the identification of unknown, though structurally similar trichothecenes, by leveraging objective ML techniques. Discriminative fragments derived from a series of gradient-boosted machine learners are then used to develop ML-driven precursor ion scan (PIS) methods on a triple quadrupole mass spectrometer (QQQ) for screening of “unknown unknown” trichothecenes. Specifically, we apply the PIS method to a laboratory-synthesized trichothecene, a first step in demonstrating the power of alternative, machine learning-driven mass spectrometric methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗