Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Generative Model”

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 1,423 records · Page 79

Application of Extended Kalman Filter Techniques for Dynamic Model Parameter Calibration

Abstract -Phasor measurement has previously been used for sub-system model validation, which enables rigorous comparison of model simulation and recorded dynamics and facilitates identification of problematic model components. Recent work extends the sub-system model validation approach with a focus on how model parameters may be calibrated to match recorded dynamics. In this paper, a calibration method using Extended Kalman Filter (EKF) technique is proposed. This paper presents the formulation as well as case studies to show the validity of the EKF-based parameter calibration method. The proposed calibration method is expected to be a cost-effective means complementary to traditional equipment testing for improving dynamic model quality.

Extended Kalman Filter↗

Modeling the Impacts of Solar Distributed Generation on U.S. Water Resources

Distributed electric power generation technologies typically use little or no water per unit of electrical energy produced; in particular, renewable energy sources such as solar PV systems do not require cooling systems and present an opportunity to reduce water usage for power generation. Within the US, the fuel mix used for power generation varies regionally, and certain areas use more water for power generation than others. The need to reduce water usage for power generation is even more urgent in view of climate change uncertainties. In this paper, we present an example case within the state of Tennessee, one of the top four states in water consumption for power generation and one of the states with little or no potential for developing centralized renewable energy generations. The potential for developing PV generation within Knox County, Tennessee, is studied, along with the potential for reducing water withdrawal and consumption within the Tennessee Valley stream region. Electric power generation plants in the region are quantified for their electricity production and expected water withdrawal and consumption over one year, where electrical generation data is provided over one year and water usage is modeled based on the cooling system(s) in use. Potential solar PV electrical production is modeled based on LiDAR data and weather data for the same year. Our proposed methodology can be summarized as follows: First, the potential solar generation is compared against the local grid demand. Next, electrical generation reductions are specified that would result in a given reduction in water withdrawal and a given reduction in water consumption, and compared with the current water withdrawal and consumption rates for the existing fuel mix. The increase in solar PV development that would produce an equivalent amount of power, is determined. In this way, we consider how targeted local actions may affect the larger stream region through thoughtful energy development. This model can be applied to other regions, other types of distributed generation, and used as a framework for modeling alternative growth scenarios in power production capacity in addition to modeling adjustments to existing capacity.

distributed energy generation↗

Laser produced plasma sources for nanolithography—Recent integrated simulation and benchmarking

Photon sources for extreme ultraviolet lithography (EUVL) are still facing challenging problems to achieve high volume manufacturing in the semiconductor industry. The requirements for high EUV power, longer optical system and components lifetime, and efficient mechanisms for target delivery have narrowed investigators towards the development and optimization of dual-pulse laser sources with high repetition rate of small liquid tin droplets and the use of multi-layer mirror optical system for collecting EUV photons. We comprehensively simulated laser-produced plasma sources in full 3D configuration using 10–50 μm tin droplet targets as single droplets as well as, for the first time, distributed fragmented microdroplets with equivalent mass. The latter is to examine the effects of droplet fragmentation resulting from the first pulse and prior to the incident second main laser pulse. We studied the dependence of target mass and size, laser parameters, and dual pulse system configuration on EUV radiation output and on atomic and ionic debris generation. Our modeling and simulation included all phases of laser target evolution: from laser/droplet interaction, energy deposition, target vaporization, ionization, plasma hydrodynamic expansion, thermal and radiation energy redistribution, and EUV photons collection as well as detail mapping of photons source size and location. We also simulated and predicted the potential damage to the optical mirror collection system from plasma thermal and energetic debris and the requirements for mitigating systems to reduce debris fluence. The debris effect on mirror collection system is analyzed using our three-dimensional ITMC-DYN Monte Carlo package. Modeling results were benchmarked against our CMUXE laboratory experimental studies for the EUV photons production and for debris and ions generation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Big Data Sanitization and Cyber Situational Awareness: A Network Telescope Perspective

This paper addresses the problems of data sanitization and cyber situational awareness by analyzing 910 GB of real Internet-scale traffic, which has been passively collected by monitoring close to 16.5 million darknet IP addresses from a /8 and a /13 network telescopes. First, the paper offers a novel probabilistic darknet preprocessing model, which aims at sanitizing darknet data to prepare it for effective use in the task of cyber threat intelligence generation. Such model has been engineered using a distributed multithreaded approach, rendering it highly effective on darknet big data. Second, the paper further contributes by presenting an innovative approach to infer large-scale orchestrated probing campaigns by leveraging darknet data, for Internet cyber situational awareness. The approach uniquely reduces the dimensionality of such big data by utilizing its artifacts, instead of processing the actual raw data. This is accomplished by extracting and analyzing probing time series using formal methods rooted in Fourier transform and Kalman filtering. Thorough empirical evaluations indeed validate the accuracy and the performance of the proposed methods. We assert that such approaches are of significant value, given their highly applicable nature to the field of Internet measurements for cyber security in the era of big data.

Bou-Harb, Elias↗

Constitutive CCND1/CDK2 Activity Substitutes for p53 Loss, or MYC or Oncogenic RAS Expression in the Transformation of Human Mammary Epithelial Cells

Cancer develops following the accumulation of genetic and epigenetic alterations that inactivate tumor suppressor genes and activate proto-oncogenes. Dysregulated cyclin-dependent kinase (CDK) activity has oncogenic potential in breast cancer due to its ability to inactivate key tumor suppressor networks and drive aberrant proliferation. Accumulation or overexpression of cyclin D1 (CCND1) occurs in a majority of breast cancers and over-expression of CCND1 leads to accumulation of activated CCND1/CDK2 complexes in breast cancer cells. We describe here the role of constitutively active CCND1/CDK2 complexes in human mammary epithelial cell (HMEC) transformation. A genetically-defined, stepwise HMEC transformation model was generated by inhibiting p16 and p53 with shRNA, and expressing exogenous MYC and mutant RAS. By replacing components of this model, we demonstrate that constitutive CCND1/CDK2 activity effectively confers anchorage independent growth by inhibiting p53 or replacing MYC or oncogenic RAS expression. These findings are consistent with several clinical observations of luminal breast cancer sub-types that show elevated CCND1 typically occurs in specimens that retain wild-type p53, do not amplify MYC, and contain no RAS mutations. Taken together, these data suggest that targeted inhibition of constitutive CCND1/CDK2 activity may enhance the effectiveness of current treatments for luminal breast cancer.

59 BASIC BIOLOGICAL SCIENCES↗

X-RAY BURST OSCILLATIONS: FROM FLAME SPREADING TO THE COOLING WAKE

Type I X-ray bursts are thermonuclear flashes observed from the surfaces of accreting neutron stars (NSs) in low mass X-ray binaries. Oscillations have been observed during the rise and/or decay of some of these X-ray bursts. Those seen during the rise can be well explained by a spreading hot spot model, but large amplitude oscillations in the decay phase remain mysterious because of the absence of a clear-cut source of asymmetry. To date there have not been any quantitative studies that consistently track the oscillation amplitude both during the rise and decay (cooling tail) of bursts. Here we compute the light curves and amplitudes of oscillations in X-ray burst models that realistically account for both flame spreading and subsequent cooling. We present results for several such “cooling wake” models, a “canonical” cooling model where each patch on the NS surface heats and cools identically, or with a latitude-dependent cooling timescale set by the local effective gravity, and an “asymmetric” model where parts of the star cool at significantly different rates. We show that while the canonical cooling models can generate oscillations in the tails of bursts, they cannot easily produce the highest observed modulation amplitudes. Alternatively, a simple phenomenological model with asymmetric cooling can achieve higher amplitudes consistent with the observations.

79 ASTRONOMY AND ASTROPHYSICS↗

Modeling the Formation of the Family of the Dwarf Planet Haumea

The dwarf planet (136108) Haumea has an intriguing combination of unique physical properties: near-breakup spin, two regular satellites, and an unexpectedly compact family. While these properties point toward formation by a collision, there is no self-consistent and reasonably probable formation hypothesis that can connect the unusually rapid spin and low relative velocities of Haumea family members (“Haumeans”). We explore and test the proposed formation hypotheses (catastrophic collision, graze-and-merge, and satellite collision) in detail. We flexibly parameterize the properties of the collision (e.g., the collision location) and use simple models for the unique three-dimensional velocity ejection field expected from each model to generate simulated families. These are then compared to the observed Kuiper Belt objects using Bayesian parameter inference, including a mixture model that robustly allows for interlopers from the background population. After testing our methodology, we find that the best match to the observed Haumeans is an essentially isotropic ejection field with a typical velocity of 150 m s{sup −1}. The graze-and-merge formation hypothesis—in which Haumeans are shed due to excess angular momentum—is clearly disfavored because the observed Haumeans are not oriented in a plane. The satellite collision model is also disfavored. Including these new constraints, we present a detailed discussion of the formation hypotheses, including variations, some of which are tested. Some new hypotheses are proposed (a cratering collision and a collision where Haumea’s upper layers are “missing”) and scrutinized. We do not identify a satisfactory formation hypothesis, but we do propose several avenues of additional investigation. In the process of these analyses, we identify many new candidate Haumeans and dynamically confirm seven of them as consistent with the observed family. We also confirm that Haumeans have a shallow size distribution and discuss implications for the discovery and identification of new Haumeans.

79 ASTRONOMY AND ASTROPHYSICS↗

System and method for assessing spectroscopic sensor accuracy

A method for assessing spectroscopic sensor accuracy, includes building an a priori simulation of generalized etalon drift. A spectroscopic sensor is tested to determine use parameters. A specific drift model is generated by applying the determined use parameters to the built a priori simulation of generalized etalon drift. The specific drift model is analyzed to determine whether the spectroscopic sensor is satisfactory.

Green, William J.↗

Automated Generation of Microkinetics for Heterogeneously Catalyzed Reactions Considering Correlated Uncertainties**

Abstract The study presents an ab‐initio based framework for the automated construction of microkinetic mechanisms considering correlated uncertainties in all energetic parameters and estimation routines. 2000 unique microkinetic models were generated within the uncertainty space of the BEEF‐vdW functional for the oxidation reactions of representative exhaust gas emissions from stoichiometric combustion engines over Pt(111) and compared to experiments through multiscale modeling. The ensemble of simulations stresses the importance of considering uncertainties. Within this set of first‐principles‐based models, it is possible to identify a microkinetic mechanism that agrees with experimental data. This mechanism can be traced back to a single exchange‐correlation functional, and it suggests that Pt(111) could be the active site for the oxidation of light hydrocarbons. The study provides a universal framework for the automated construction of reaction mechanisms with correlated uncertainty quantification, enabling a DFT‐constrained microkinetic model optimization for other heterogeneously catalyzed systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Neural Posterior Estimation for Scalable and Accurate Inverse Parameter Inference in Li-Ion Batteries

Diagnosing the internal state of Li-ion batteries is critical for battery research, operation of real-world systems, and prognostic evaluation of remaining lifetime. By using physics-based models to perform probabilistic parameter estimation via Bayesian calibration, diagnostics can account for the uncertainty due to model fitness, data noise, and the observability of any given parameter. However, Bayesian calibration in Li-ion batteries using electrochemical data is computationally intensive even when using a fast surrogate in place of physics-based models, requiring many thousands of model evaluations. A fully amortized alternative is neural posterior estimation (NPE). NPE shifts the computational burden from the parameter estimation step to data generation and model training, reducing the parameter estimation time from minutes to milliseconds, enabling real-time applications. The present work shows that NPE can infer parameters equally or more accurately than Bayesian calibration, even if it leads to higher voltage reconstruction errors. We also demonstrate that the higher computational costs for data generation are tractable even in high-dimensional cases (ranging from 6 to 27 estimated parameters). The NPE method also offers several interpretability advantages over Bayesian calibration, such as local parameter sensitivity to specific regions of the voltage curve. The NPE method is demonstrated using an experimental fast charge dataset, with parameter estimates validated against measurements of loss of lithium inventory and loss of active material. The implementation is made available in a companion repository (https://github.com/NatLabRockies/BatFIT).

25 ENERGY STORAGE↗

A Process for Defining Prototype Building Models: Courthouse Case Study for U.S. Commercial Energy

Buildings currently consume 36% of the world’s energy and contribute nearly 40% of CO 2 emissions. Many countries desire to generate virtual models of their nation’s buildings in order to coordinate research activities and inform market opportunities for a more sustainable built environment. The United States Department of Energy uses a suite of Commercial Prototype Building Models, which currently includes 16 building types and covers 80% of US commercial floorspace. Efforts are underway to expand this suite by developing prototype models for additional building types. In this paper, we outline a systematic approach to defining the building, collecting relevant information and creating a flexible model while doing so in the pragmatic context of a courthouse building. Informed by building design guides, databases, documented projects and inputs from courthouse design experts, we define a small, 69,324 ft 2 (6440 m 2 ), four-courtroom, low-rise courthouse as the prototype to represent an average-size courthouse in the US. We present building characteristics relevant for energy model development and provide the rationale for their selection. These details combined with climate- and construction-vintage-specific requirements for the building envelope and systems from building standards will be used for developing the courthouse model for the Commercial Prototype Building Models suite. The comprehensive information presented will also guide model modification to capture the dynamics of smaller or larger courthouses more accurately for building or system size-specific research.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Minnesota Solar Pathways: Illuminating Pathways to 10% Solar (Final Technical Report)

The Minnesota Solar Pathways project, funded in part by the Department of Energy’s Solar Energy Technologies Office, was a three-year project designed to explore least-risk, best-value strategies for meeting the State of Minnesota’s solar goals. As part of this aim, the Pathways Team modeled renewable generation costs, examined ways to streamline interconnection, and evaluated technologies that can increase solar hosting capacity on the distribution grid. Midway through the project period, the Pathways Team expanded the scope of study using the same modeling framework from the Minnesota Solar Potential Analysis to estimate and compare generation costs for high renewables penetration scenarios across the Midcontinent Independent System Operator (MISO) region. The analysis concludes that it can be cheaper to overbuild solar + wind and curtail surplus than to store all generation. In addition, the analysis of individual MISO subregions compared with the whole region highlights the opportunities of enabling market potential compared with impact of ongoing transmission constraints. Beyond the analysis of future scenarios, the partnerships initiated and supported through this project underscored the ongoing need for work with stakeholders to address key issues, particularly around siting, and leverage symbiotic strategies to meet state goals.

14 SOLAR ENERGY↗

Predictive model for real-time energy disaggregation using long short-term memory

To provide affordable energy-saving solutions for the small and medium-sized manufacturers (SMMs), we propose a unified framework for generating predictive models that support real-time disaggregation of power consumption from combined inputs, enabling automatic machine state identification simultaneously for joint analysis of energy usage patterns. Further, the proposed framework transforms raw power consumption into a time series with look-back and bootstrap capabilities for historical pattern detection, while a learning architecture utilizes the stacked long short-term memory (LSTM) layers as encoders for embedding generation with sequential awareness. Experimental results demonstrate 93.65% minimum accuracy in ideal case of real-time energy usage and machine state prediction.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Models For Laser Ablation Mass Removal And Impulse Generation In Vacuum

To the present day, literature efforts at modeling laser propulsion impulse often used empirical models. Recently, a simple physical approach was demonstrated to be effective for predicting many practical properties of laser ablative impulse generation under vacuum. The model used photochemical mass removal and energy conservation to predict parameters such as the peak momentum coupling coefficient, the optimal fluence position at which this maximum is reached, and various critical properties related to the laser ablation threshold. Although the current model understanding is not complete, improvements in the treatment of mass removal and ambient pressure are expected to allow this type of model to be broadly applicable to many diverse applications using laser ablation impulse generation. In this paper, we also introduce an alternative formulation of the model incorporating photothermal mass removal. Implications and limitations of the model formulation in its initial stage of development are discussed, particularly concerning critical fluence effects and directions for improvement.

42 ENGINEERING↗

Improving Tumor Treating Fields Treatment Efficacy in Patients With Glioblastoma Using Personalized Array Layouts

Purpose: To investigate tumors of different size, shape, and location and the effect of varying transducer layouts on Tumor Treating Fields (TTFields) distribution in an anisotropic model. Methods and Materials: A realistic human head model was generated from MR images of 1 healthy subject. Four different virtual tumors were placed at separate locations. The transducer arrays were modeled to mimic the TTFields-delivering commercial device. For each tumor location, varying array layouts were tested. The finite element method was used to calculate the electric field distribution, taking into account tissue heterogeneity and anisotropy. Results: In all tumors, the average electric field induced by either of the 2 perpendicular array layouts exceeded the 1-V/cm therapeutic threshold value for TTFields effectiveness. Field strength within a tumor did not correlate with its size and shape but was higher in more superficial tumors. Additionally, it always increased when the array was adapted to the tumor's location. Compared with a default layout, the largest increase in field strength was 184%, and the highest average field strength induced in a tumor was 2.21 V/cm. Conclusions: These results suggest that adapting array layouts to specific tumor locations can significantly increase field strength within the tumor. Our findings support the idea of personalized treatment planning to increase TTFields efficacy for patients with GBM.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

No electrostatic supersolitons in two-component plasmas

The concept of acoustic supersolitons was introduced for a very specific plasma with five constituents, and discussed only for a single set of plasma parameters. Supersolitons are characterized by having subsidiary extrema on the sides of a typical bipolar electric field signature, or by association with a root beyond double layers in the fully nonlinear Sagdeev pseudopotential description. It was subsequently found that supersolitons could exist in several plasma models having three constituent species, rather than four or five. In the present paper, it is proved that standard two-component plasma models cannot generate supersolitons, by recalling and extending results already in the literature, and by establishing the necessary properties of a more recent model.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗