Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “robustness”

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 1,369 records · Page 76
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Distributionally Robust Building Load Control to Compensate Fluctuations in Solar Power Generation

This paper investigates the use of a collection of dispatchable heating, ventilation and air conditioning (HVAC) systems to absorb low-frequency fluctuations in renewable energy sources, especially in solar photo-voltaic (PV) generation. Given the uncertain and time-varying nature of solar PV generation, its probability distribution is difficult to be estimated perfectly, which poses a challenging problem of how to optimally schedule a fleet of HVAC loads to consume as much as local PV generation. We formulate a distributionally robust chance-constrained (DRCC) model to ensure that PV generation is consumed with a desired probability for a family of probability distributions, termed as an ambiguity set, built upon mean and covariance information. We benchmark the DRCC model with a deterministic optimization model and a stochastic programming model in a one-day simulation. We show that the DRCC model achieves constantly good performance to consume most PV generation even in the case with the presence of probability distribution ambiguity.

Zhang, Yiling↗
DOE OSTI2022

Distributed Transient Safety Verification via Robust Control Invariant Sets: A Microgrid Application

Modern safety-critical energy infrastructures are increasingly operated in a hierarchical and modular control framework which allows for limited data exchange between the modules. In this context, it is important for each module to synthesize and communicate constraints on the values of exchanged information in order to assure system-wide safety. To ensure transient safety in inverter-based microgrids, we develop a set invariance-based distributed safety verification algorithm for each inverter module. Applying Nagumo's invariance condition, we construct a robust polynomial optimization problem to jointly search for safety-admissible set of control set-points and design parameters, under allowable disturbances from neighbors. We use sum-of-squares (SOS) programming to solve the verification problem and we perform numerical simulations using grid-forming inverters to illustrate the algorithm.

Bouvier, Jean-Baptiste H.↗
DOE OSTI2023

Robust Decentralized Secondary Control Scheme for Inverter-based Power Networks

Inverter-dominated microgrids are quickly becoming a key building block of future power systems. They rely on centralized controllers that can provide reliability and resiliency in extreme events. Nonetheless, communication failures due to cyber-physical attacks or natural disasters can make autonomous operation of islanded microgrids challenging. This paper examines a unified decentralized secondary control scheme that is robust to inverter clock synchronization errors and can be seamlessly applied to grid-following or grid-forming control architectures. The proposed scheme overcomes the well-known stability problem that arises from parallel operation of local integral controllers. Theoretical guarantees for stability are provided along with criteria to appropriately tune the secondary control gains to achieve good frequency regulation performance while ensuring fair power sharing. The efficacy of our approach is demonstrated through simulations on a 5-bus microgrid with four grid-forming inverters.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2024

ExaWorks software development kit: a robust and scalable collection of interoperable workflows technologies

Scientific discovery increasingly requires executing heterogeneous scientific workflows on high-performance computing (HPC) platforms. Heterogeneous workflows contain different types of tasks (e.g., simulation, analysis, and learning) that need to be mapped, scheduled, and launched on different computing. That requires a software stack that enables users to code their workflows and automate resource management and workflow execution. Currently, there are many workflow technologies with diverse levels of robustness and capabilities, and users face difficult choices of software that can effectively and efficiently support their use cases on HPC machines, especially when considering the latest exascale platforms. We contributed to addressing this issue by developing the ExaWorks Software Development Kit (SDK). The SDK is a curated collection of workflow technologies engineered following current best practices and specifically designed to work on HPC platforms. We present our experience with (1) curating those technologies, (2) integrating them to provide users with new capabilities, (3) developing a continuous integration platform to test the SDK on DOE HPC platforms, (4) designing a dashboard to publish the results of those tests, and (5) devising an innovative documentation platform to help users to use those technologies. Our experience details the requirements and the best practices needed to curate workflow technologies, and it also serves as a blueprint for the capabilities and services that DOE will have to offer to support a variety of scientific heterogeneous workflows on the newly available exascale HPC platforms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation

The quantitative characterization of surface structures captured in scanning electron microscopy (SEM) images has proven to be effective for discerning provenance of an unknown nuclear material. Recently, many works have taken advantage of the powerful performance of convolutional neural networks (CNNs) to provide faster and more consistent characterization of surface structures. However, one inherent limitation of CNNs is their degradation in performance when encountering discrepancy between training and test datasets, which limits their use widely. The common discrepancy in an SEM image dataset occurs at low-level image information due to user-bias in selecting acquisition parameters and microscopes from different manufacturers. Therefore, in this study, we present a domain adaptation framework to improve robustness of CNNs against the discrepancy in low-level image information. Furthermore, our proposed approach makes use of only unlabeled test samples to adapt a pretrained model, which is more suitable for nuclear forensics application for which obtaining both training and test datasets simultaneously is a challenge due to data sensitivity. Through extensive experiments, we demonstrate that our proposed approach effectively improves the performance of a model by at least 18% when encountering domain discrepancy, and can be deployed in many CNN architectures.

scanning electron microscopy↗
DOE OSTI2021

Fluorescent Detection of Carbon Disulfide by a Highly Emissive and Robust Isoreticular Series of Zr-Based Luminescent Metal Organic Frameworks (LMOFs)

Carbon disulfide (CS 2 ) is a highly volatile neurotoxic species. It is known to cause atherosclerosis and coronary artery disease and contributes significantly to sulfur-based pollutants. Therefore, effective detection and capture of carbon disulfide represents an important aspect of research efforts for the protection of human and environmental health. In this study, we report the synthesis and characterization of two strongly luminescent and robust isoreticular metal organic frameworks (MOFs) Zr 6 (µ 3- O) 4 (OH) 8 (tcbpe) 2 (H 2 O) 4 (here termed 1) and Zr 6 (µ 3- O) 4 (OH) 8 (tcbpe-f) 2 (H 2 O) 4 (here termed 2) and their use as fluorescent sensors for the detection of carbon disulfide. Both MOFs demonstrate a calorimetric bathochromic shift in the optical bandgap and strong luminescence quenching upon exposure to carbon disulfide. The interactions between carbon disulfide and the frameworks are analyzed by in-situ infrared spectroscopy and computational modelling by density functional theory. These results reveal that both the Zr metal node and organic ligand act as the preferential binding sites and interact strongly with carbon disulfide.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Planning Amidst Uncertainty: Identifying Core CCS Infrastructure Robust to Storage Uncertainty

Carbon Capture and Storage (CCS) is a critical technology for reducing anthropogenic CO2 emissions, but its large-scale deployment is complicated by uncertainties in geological storage performance. These uncertainties pose significant financial and operational risks, as underperforming storage sites can lead to costly infrastructure modifications, inefficient pipeline routing, and economic shortfalls. To address this challenge, we propose a novel optimization workflow that is based on mixed-integer linear programming and explicitly integrates probabilistic modeling of storage uncertainty into CCS infrastructure design. This workflow generates multiple infrastructure scenarios by sampling storage capacity distributions, optimally solving each scenario using a mixed-integer linear programming model, and aggregating results into a heatmap to identify core infrastructure components that have a low likelihood of underperforming. A risk index parameter is introduced to balance trade-offs between cost, CO2 processing capacity, and risk of underperformance, allowing stakeholders to quantify and mitigate uncertainty in CCS planning. Applying this workflow to a CCS dataset from the US Department of Energy’s Carbon Utilization and Storage Partnership project reveals key insights into infrastructure resilience. Reducing the risk index from 15% to 0% is observed to lead to an 83.7% reduction in CO2 processing capacity and a 77.1% decrease in project profit, quantifying the trade-off between risk tolerance and project performance. Furthermore, our results highlight critical breakpoints, where small adjustments in the risk index produce disproportionate shifts in infrastructure performance, providing actionable guidance for decision-makers. Unlike prior approaches that aimed to cheaply repair underperforming infrastructure, our workflow constructs robust CCS networks from the ground up, ensuring cost-effective infrastructure under storage uncertainty. These findings demonstrate the practical relevance of incorporating uncertainty-aware optimization into CCS planning, equipping decision-makers with a tool to make informed project planning decisions.

Olson, Daniel↗
DOE OSTI2020

Robust Learning with Implicit Residual Networks

In this effort, we propose a new deep architecture utilizing residual blocks inspired by implicit discretization schemes. As opposed to the standard feed-forward networks, the outputs of the proposed implicit residual blocks are defined as the fixed points of the appropriately chosen nonlinear transformations. We show that this choice leads to the improved stability of both forward and backward propagations, has a favorable impact on the generalization power, and allows for control the robustness of the network with only a few hyperparameters. In addition, the proposed reformulation of ResNet does not introduce new parameters and can potentially lead to a reduction in the number of required layers due to improved forward stability. Finally, we derive the memory-efficient training algorithm, propose a stochastic regularization technique, and provide numerical results in support of our findings.

Reshniak, Viktor↗
DOE OSTI2019

On a Robust, Sensitive Cell-Free Method for Pseudomonas Sensing and Quantification in Microfluidic Templated Hydrogels

Through the use of droplet microfluidics to integrate cell-free activity into inert hydrogel beads, we have developed a platform that can perform biologically relevant functions without the need for cells. Specifically, cell-free lysates serve a utility in performing cellular functions and providing biologically relevant metabolic products without requiring the optimal biological conditions for cell growth and proliferation. By teasing out specific biological components that enable transcription and translation to occur, these cell-like functions can be reconstituted in vitro without requiring the entire cell and milieu of cellular organelles. This enables the optimization of synthetic biological circuits, either by concentration or logic switches, simply through the addition or removal of genetic components (plasmids, inducers, or repressors) of regulatory elements. Here, we demonstrate an application of cell-free processes that is robust and portable, independent of a substrate, to apply for sensing and reporting functions of a quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12HSL) found crucial for pathological Pseudomonas aeruginosa infection. We develop an agarose bead platform that is easily adaptable and simply programmable to fit a variety of biological and chemical sensing applications for the utility of ease of delivery and activation in remote environments—even in conditions with very little hydration.

Instruments & Instrumentation↗
DOE OSTI2020

A Robust Segmented Mixed Effect Regression Model for Baseline Electricity Consumption Forecasting

Renewable energy production has been surging around the world in recent years. To mitigate the increasing uncertainty and intermittency of the renewable generation, proactive demand response algorithms and programs are proposed and developed to further improve the utilization of load flexibility and increase the efficiency of power system operation. One of the biggest challenges to efficient control and operation of demand response resources is how to forecast the baseline electricity consumption and estimate the load impact from demand response resources accurately. In this paper, we propose a mixed effect segmented regression model and a new robust estimate for forecasting the baseline electricity consumption in Southern California, USA, by combining the ideas of random effect regression model, segmented regression model, and the least trimmed squares estimate. Since the log-likelihood of the considered model is not differentiable at breakpoints, we propose a new backfitting algorithm to estimate the unknown parameters. The estimation performance of the new estimation procedure has been demonstrated with both simulation studies and the real data application for the electric load baseline forecasting in Southern California.

42 ENGINEERING↗
DOE OSTI2016

On the robustness of the pendulum model for large-amplitude longitudinal oscillations in prominences

Large-amplitude longitudinal oscillations (LALOs) in prominences are spectacular manifestations of solar activity. In such events nearby energetic disturbances induce periodic motions on filaments with displacements comparable to the size of the filaments themselves and with velocities larger than 20 km s{sup −1}. The pendulum model, in which the gravity projected along a rigid magnetic field is the restoring force, was proposed to explain these events. However, it can be objected that in a realistic situation where the magnetic field reacts to the mass motion of the heavy prominence, the simplified pendulum model could be no longer valid. We have performed nonlinear time-dependent numerical simulations of LALOs considering a dipped magnetic field line structure. In this work we demonstrate that for even relatively weak magnetic fields the pendulum model works very well. We therefore validate the pendulum model and show its robustness, with important implications for prominence seismology purposes. With this model it is possible to infer the geometry of the dipped field lines that support the prominence.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Data assimilation in a solar dynamo model using Ensemble Kalman Filters: sensitivity and robustness in reconstruction of meridional flow speed

We implement an Ensemble Kalman Filter procedure using the Data Assimilation Research Testbed for assimilating “synthetic” meridional flow-speed data in a Babcock–Leighton-type flux-transport solar dynamo model. By performing several “observing system simulation experiments,” we reconstruct time variation in meridional flow speed and analyze sensitivity and robustness of reconstruction. Using 192 ensemble members including 10 observations, each with 4% error, we find that flow speed is reconstructed best if observations of near-surface poloidal fields from low latitudes and tachocline toroidal fields from midlatitudes are assimilated. If observations include a mixture of poloidal and toroidal fields from different latitude locations, reconstruction is reasonably good for ⩽40% error in low-latitude data, even if observational error in polar region data becomes 200%, but deteriorates when observational error increases in low- and midlatitude data. Solar polar region observations are known to contain larger errors than those in low latitudes; our forward operator (a flux-transport dynamo model here) can sustain larger errors in polar region data, but is more sensitive to errors in low-latitude data. An optimal reconstruction is obtained if an assimilation interval of 15 days is used; 10- and 20-day assimilation intervals also give reasonably good results. Assimilation intervals <5 days do not produce faithful reconstructions of flow speed, because the system requires a minimum time to develop dynamics to respond to flow variations. Reconstruction also deteriorates if an assimilation interval >45 days is used, because the system’s inherent memory interferes with its short-term dynamics during a substantially long run without updating.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

A robust derivation of the tight relationship of radio core dominance to inclination angle in high redshift 3CRR sources

It is believed that, in radio-loud active galactic nuclei (AGNs), the core radio flux density can be normalized to the flux density of the extended lobe emission to infer the orientation of a radio source. However, very little is known about the reliability and precision of this method, and we are unaware of any robust conversion recipe to infer the inclination from the core dominance. Investigating whether or not the radio core dominance parameter R separates the quasars from the radio galaxies in the z ⩾ 1 3CRR catalog, we found excellent agreement of R with optical type, infrared flux ratios, and optical polarization. This indicates that probably both R and optical classification are very good orientation indicators, and the unified model is strongly predictive for these objects. The relative number densities indicate half-opening angles close to 60°, as expected from large surveys. The separations of optical types according to radio core dominance as well as near-infrared/MIR ratios, which are essentially perfect, mean that there can be only a small dispersion of torus half-opening angles. Also, even though torus dust is thought to be clumpy, there is an almost zero probability to see a type-1 source at high inclination. Finally, using only the Copernican Principle, i.e., the assumption that solid angle is filled uniformly with source axis orientations, we estimated a semi-empirical relation between core dominance and AGN inclination. This makes it possible to use R to infer the inclination of a source to an accuracy of ∼10° or less, at least for this type of object.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

21 cm Signal Recovery via Robust Principal Component Analysis

The redshifted 21 cm signal from neutral hydrogen (H i) is potentially a very powerful probe for cosmology, but a difficulty in its observation is that it is much weaker than foreground radiation from the Milky Way as well as extragalactic radio sources. The foreground radiation at different frequencies, however, is coherent along one line of sight, and various methods of foreground subtraction based on this property have been proposed. In this paper, we present a new method based on robust principal component analysis (RPCA) to subtract the foreground and extract the 21 cm signal, which explicitly uses both the low-rank property of the frequency covariance matrix (i.e., frequency coherence) of the foreground and the sparsity of the frequency covariance matrix of the 21 cm signal. The low-rank property of the foreground’s frequency covariance has been exploited in many previous works on foreground subtraction, but to our knowledge the sparsity of the frequency covariance of the 21 cm signal is first explored here. By exploiting both properties in the RPCA method, the frequency covariance matrix of the foreground and the 21 cm signal can be separated accurately. We then use the generalized internal linear combination method to recover the 21 cm signal from the data. Our method is applicable both to a small patch of sky with the flat-sky approximation and to a large area of sky where the sphericity has to be considered. It is also easy to extend to deal with more complex conditions such as a sky map with defects.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

The 2015 decay of the black hole X-ray binary V404 Cygni: robust disk-jet coupling and a sharp transition into quiescence

We present simultaneous X-ray and radio observations of the black hole X-ray binary V404 Cygni at the end of its 2015 outburst. From 2015 July 11–August 5, we monitored V404 Cygni with Chandra, Swift, and NuSTAR in the X-ray, and with the Karl G. Jansky Very Large Array and the Very Long Baseline Array in the radio, spanning a range of luminosities that were poorly covered during its previous outburst in 1989 (our 2015 campaign covers 2×10{sup 33}≲L{sub X}≲10{sup 34} erg s{sup −1}). During our 2015 campaign, the X-ray spectrum evolved rapidly from a hard photon index of Γ≈1.6 (at L{sub X}≈10{sup 34} erg s{sup −1}) to a softer Γ≈2 (at L{sub X}≈3×10{sup 33} erg s{sup −1}). We argue that V404 Cygni reaching Γ≈2 marks the beginning of the quiescent spectral state, which occurs at a factor of ≈3–4 higher X-ray luminosity than the average pre-outburst luminosity of ≈8×10{sup 32} erg s{sup −1}. V404 Cygni falls along the same radio/X-ray luminosity correlation that it followed during its previous outburst in 1989, implying a robust disk-jet coupling. We exclude the possibility that a synchrotron-cooled jet dominates the X-ray emission in quiescence, leaving synchrotron self-Compton from either a hot accretion flow or from a radiatively cooled jet as the most likely sources of X-ray radiation, and/or particle acceleration along the jet becoming less efficient in quiescence. Finally, we present the first indications of correlated radio and X-ray variability on minute timescales in quiescence, tentatively measuring the radio emission to lag the X-ray by 15±4 minute, suggestive of X-ray variations propagating down a jet with a length of <3.0 au.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Advection of Potential Temperature in the Atmosphere of Irradiated Exoplanets: A Robust Mechanism to Explain Radius Inflation

The anomalously large radii of strongly irradiated exoplanets have remained a major puzzle in astronomy. Based on a two-dimensional steady-state atmospheric circulation model, the validity of which is assessed by comparison to three-dimensional calculations, we reveal a new mechanism, namely the advection of the potential temperature due to mass and longitudinal momentum conservation, a process occurring in the Earth’s atmosphere or oceans. In the deep atmosphere, the vanishing heating flux forces the atmospheric structure to converge to a hotter adiabat than the one obtained with 1D calculations, implying a larger radius for the planet. Not only do the calculations reproduce the observed radius of HD 209458b, but also reproduce the observed correlation between radius inflation and irradiation for transiting planets. Vertical advection of potential temperature induced by non-uniform atmospheric heating thus provides a robust mechanism to explain the inflated radii of irradiated hot Jupiters.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Robust Cross-correlation-based Measurement of Clump Sizes in Galaxies

Stars form in molecular complexes that are visible as giant clouds (∼10{sup 5−6} M{sub ⊙}) in nearby galaxies and as giant clumps (∼10{sup 8−9} M{sub ⊙}) in galaxies at redshifts z≈1–3. Theoretical inferences on the origin and evolution of these complexes often require robust measurements of their characteristic size, which is hard to measure at limited resolution and often ill-defined due to overlap and quasi-fractal substructure. We show that maximum and luminosity-weighted sizes of clumps seen in star formation maps (e.g., Hα) can be recovered statistically using the two-point correlation function (2PCF) if an approximate stellar surface density map is taken as the normalizing random field. After clarifying the link between Gaussian clumps and the 2PCF analytically, we design a method for measuring the diameters of Gaussian clumps with realistic quasi-fractal substructure. This method is tested using mock images of clumpy disk galaxies at different spatial resolutions and perturbed by Gaussian white noise. We find that the 2PCF can recover the input clump scale at ∼20% accuracy, as long as this scale is larger than the spatial resolution. We apply this method to the local spiral galaxy NGC 5194, as well as to three clumpy turbulent galaxies from the DYNAMO-HST sample. In both cases, our statistical measurements of Hα clump size agree with previous measurements and with the estimated Jeans lengths. However, the new measurements are free from subjective choices when fitting individual clumps.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

A Robust and Efficient Deep Learning Method for Dynamical Mass Measurements of Galaxy Clusters

We demonstrate the ability of convolutional neural networks (CNNs) to mitigate systematics in the virial scaling relation and produce dynamical mass estimates of galaxy clusters with remarkably low bias and scatter. We present two models, CNN 1D and CNN 2D , which leverage this deep learning tool to infer cluster masses from distributions of member galaxy dynamics. Our first model, CNN 1D , infers cluster mass directly from the distribution of member galaxy line-of-sight velocities. Our second model, CNN 2D , extends the input space ofCNN 1D to learn on the joint distribution of galaxy line-of-sight velocities and projected radial distances. We train each model as a regression over cluster mass using a labeled catalog of realistic mock cluster observations generated from the MultiDark simulation and UniverseMachine catalog. We then evaluate the performance of each model on an independent set of mock observations selected from the same simulated catalog. The CNN models produce cluster mass predictions with lognormal residuals of scatter as low as 0.132 dex, greater than a factor of 2 improvement over the classical M–σ power-law estimator. Furthermore, the CNN model reduces prediction scatter relative to similar machinelearning approaches by up to 17% while executing in drastically shorter training and evaluation times (by a factor of 30) and producing considerably more robust mass predictions (improving prediction stability under variations in galaxy sampling rate by 30%).

79 ASTRONOMY AND ASTROPHYSICS↗