Search NASA⌕ Search

SEARCH · Search NASA

Results for “robust estimation”

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 739 records · Page 41

Solution-Based Adaptivity as a Paradigm for Computational Fluid Dynamics

This is the final report for the DOE Early Career award. Through innovation in error estimation and solution-based adaptation methods we strive to dramatically improve the efficiency and robustness of computational science and engineering simulations. By solution-based error estimation and adaptation we refer to techniques for estimating numerical error and efficiently reducing it through targeted discretization changes. In such a setting, high-order methods play a key role, though they are just one mechanism for efficient accuracy improvement. Other mechanisms include well known mesh size and stretching optimization, as well as less-practiced mesh motion and basis optimization. With additional work on the latter, we are integrating these techniques into one adaptive suite that maximizes efficiency and robustness of Computational Fluid Dynamics (CFD) calculations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Lateral control system design for VTOL landing on a DD963 in high sea states

The problem of designing lateral control systems for the safe landing of VTOL aircraft on small ships is addressed. A ship model is derived. The issues of estimation and prediction of ship motions are discussed, using optimal linear linear estimation techniques. The roll motion is the most important of the lateral motions, and it is found that it can be predicted for up to 10 seconds in perfect conditions. The automatic landing of the VTOL aircraft is considered, and a lateral controller, defined as a ship motion tracker, is designed, using optimal control techniqes. The tradeoffs between the tracking errors and the control authority are obtained. The important couplings between the lateral motions and controls are demonstrated, and it is shown that the adverse couplings between the sway and the roll motion at the landing pad are significant constraints in the tracking of the lateral ship motions. The robustness of the control system, including the optimal estimator, is studied, using the singular values analysis. Through a robustification procedure, a robust control system is obtained, and the usefulness of the singular values to define stability margins that take into account general types of unstructured modelling errors is demonstrated. The minimal destabilizing perturbations indicated by the singular values analysis are interpreted and related to the multivariable Nyquist diagrams.

Bodson, M.↗

Parameter Estimation for Electrode Degradation: Learning in the Face of Model-Experiment Discrepancies

Use of physics-based models to interpret battery degradation data over the course of cycling can provide deeper physical insight into the internal states of the system and how they evolve. We present a neural network trained on simulations generated by a previously published physics-based model for a lithium trivanadate (LVO) cathode to estimate parameters that evolve over the course of cycling. We focus on the robustness of the neural network through two case studies that probe different kinds of discrepancies between model and experiment: nonideal data and imperfect model. In the former, the experimental protocols do not meet the assumption made in the training data generated by the physics-based model, while in the latter, the physics-based model fails to describe all of the measured cathode behavior even under ideal conditions. When there is total model-experiment agreement, a neural network estimates parameters with improved accuracy compared to a maximum likelihood analysis using the same set of simulations. However, in both types of model-experiment discrepancy, the neural network returned biased parameter estimates. We introduce a data augmentation procedure into the neural network training to mitigate these effects and improve robustness, and employ it to estimate parameters for a cycling LVO cathode.

Mayilvahanan, Karthik S. (ORCID:0000000316282332)↗

TH-CD-209-10: Scanning Proton Arc Therapy (SPArc) - The First Robust and Delivery-Efficient Spot Scanning Proton Arc Therapy

Purpose: To develop a delivery-efficient proton spot-scanning arc therapy technique with robust plan quality. Methods: We developed a Scanning Proton Arc(SPArc) optimization algorithm integrated with (1)Control point re-sampling by splitting control point into adjacent sub-control points; (2)Energy layer re-distribution by assigning the original energy layers to the new sub-control points; (3)Energy layer filtration by deleting low MU weighting energy layers; (4)Energy layer re-sampling by sampling additional layers to ensure the optimal solution. A bilateral head and neck oropharynx case and a non-mobile lung target case were tested. Plan quality and total estimated delivery time were compared to original robust optimized multi-field step-and-shoot arc plan without SPArc optimization (Arcmulti-field) and standard robust optimized Intensity Modulated Proton Therapy(IMPT) plans. Dose-Volume-Histograms (DVH) of target and Organ-at-Risks (OARs) were analyzed along with all worst case scenarios. Total delivery time was calculated based on the assumption of a 360 degree gantry room with 1 RPM rotation speed, 2ms spot switching time, beam current 1nA, minimum spot weighting 0.01 MU, energy-layer-switching-time (ELST) from 0.5 to 4s. Results: Compared to IMPT, SPArc delivered less integral dose(−14% lung and −8% oropharynx). For lung case, SPArc reduced 60% of skin max dose, 35% of rib max dose and 15% of lung mean dose. Conformity Index is improved from 7.6(IMPT) to 4.0(SPArc). Compared to Arcmulti-field, SPArc reduced number of energy layers by 61%(276 layers in lung) and 80%(1008 layers in oropharynx) while kept the same robust plan quality. With ELST from 0.5s to 4s, it reduced 55%–60% of Arcmulti-field delivery time for the lung case and 56%–67% for the oropharynx case. Conclusion: SPArc is the first robust and delivery-efficient proton spot-scanning arc therapy technique which could be implemented in routine clinic. For modern proton machine with ELST close to 0.5s, SPArc would be a popular treatment option for both single and multi-room center.

60 APPLIED LIFE SCIENCES↗

Modeling clear-air turbulence with vortices using parameter-identification techniques

A vortex model of winds associated with clear-air turbulence (CAT) is shown to be useful for characterizing actual airline CAT encounters. The model consists of an array of vortices with solid-body cores embedded in a potential flow field. Parameters such as the size and strength of the vortices and their locations are identified using a modified Newton-Raphson algorithm. A manual identification start-up scheme is used to minimize errors in the initial parameter estimates, and the identification algorithm is found to be robust in regard to the remaining errors. The analysis of a CAT encounter involving a commercial airliner demonstrates the success of the model and estimation procedure. The analysis finds vortices with core diameters of 1000 ft and tangential velocities of 87 ft/sec in this encounter.

Mehta, R. S.↗

Brownian bridge-based speed imputation technique for truck energy consumption and emissions estimation

The available truck Global Positioning System (GPS) data, typically collected with large time gaps, rely on imputation techniques to obtain second-by-second data that are required in models for estimating truck energy consumption and emissions. However, existing speed imputation techniques either require a large amount of high-resolution data for model training or rely on special movement assumptions. Here, to fill the gap and effectively apply the low-resolution truck GPS datasets, this paper proposes a simple imputation technique that adopts the Brownian bridge structure to impute missing speed data. The proposed technique introduces a feasible imputation region and a combined drift into the imputation procedure to capture vehicle acceleration constraint, travel distance constraint, and speed volatility. The calibrated model is applied to a set of low-resolution truck GPS data. The results demonstrate the robustness of the proposed technique in enhancing estimation accuracy when using low-resolution GPS data to estimate fuel consumption and emissions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

ReLU, Sparseness, and the Encoding of Optic Flow in Neural Networks

Accurate self-motion estimation is critical for various navigational tasks in mobile robotics. Optic flow provides a means to estimate self-motion using a camera sensor and is particularly valuable in GPS- and radio-denied environments. The present study investigates the influence of different activation functions—ReLU, leaky ReLU, GELU, and Mish—on the accuracy, robustness, and encoding properties of convolutional neural networks (CNNs) and multi-layer perceptrons (MLPs) trained to estimate self-motion from optic flow. Our results demonstrate that networks with ReLU and leaky ReLU activation functions not only achieved superior accuracy in self-motion estimation from novel optic flow patterns but also exhibited greater robustness under challenging conditions. The advantages offered by ReLU and leaky ReLU may stem from their ability to induce sparser representations than GELU and Mish do. Our work characterizes the encoding of optic flow in neural networks and highlights how the sparseness induced by ReLU may enhance robust and accurate self-motion estimation from optic flow.

97 MATHEMATICS AND COMPUTING↗

Identification of robust microsatellite markers for wild pig fecal DNA

ABSTRACT Collection of fecal samples for use in a genetic capture‐mark‐recapture framework has become popular as a noninvasive method of monitoring wildlife populations. A major caveat to this process, however, is that fecal samples often yield low quality DNA that is prone to genotyping errors, potentially leading to biases in population parameter estimation. Therefore, considerable care is required to identify robust genetic markers, especially in hot or humid conditions that may accelerate DNA degradation. We identified microsatellite loci in wild pig ( Sus scrofa ) fecal samples that were robust and informative within warm, humid ecosystems. To examine how degradation affected genotyping success, we sampled pig feces across 5 days and calculated how the number of quantitative polymerase chain reaction (qPCR) cycles required to reach the fluorescent threshold ( C t ) changed over time. We identified 17 microsatellite loci that had high polymorphism and amplification success and low genotyping error rates (0–0.050 per locus). In the degradation experiment, C t increased over the 5 days, but in the absence of rain, the majority of samples produced accurate genotypes after 5 days (2,211/2,550 genotypes). Based on the high amplification success and low error rates, even after 5 days of exposure to warm, humid conditions, these loci are useful for estimating population parameters in pig fecal samples. © 2016 The Wildlife Society.

Kierepka, Elizabeth M.↗

Kinetic parameters of the GUINEVERE reference configuration in VENUS-F reactor obtained from a pile noise experiment using Rossi and Feynman methods

A pile noise measurement campaign has been conducted by the CEA in the VENUS-F reactor (SCK-CEN, Mol Belgium) in April 2011 in the reference critical configuration of the GUINEVERE experimental program. The experimental setup made it possible to estimate the core kinetic parameters: the prompt neutron decay constant, the delayed neutron fraction and the generation time. A precise assessment of these constants is of prime importance. In particular, the effective delayed neutron fraction is used to normalize and compare calculated reactivities of different subcritical configurations, obtained by modifying either the core layout or the control rods position, with experimental ones deduced from the analysis of measurements. This paper presents results obtained with a CEA-developed time stamping acquisition system. Data were analyzed using Rossi-α and Feynman-α methods. Results were normalized to reactor power using a calibrated fission chamber with a deposit of Np-237. Calculated factors were necessary to the analysis: the Diven factor was computed by the ENEA (Italy) and the power calibration factor by the CNRS/IN2P3/LPC Caen. Results deduced with both methods are consistent with respect to calculated quantities. Recommended values are given by the Rossi-α estimator, that was found to be the most robust. The neutron generation time was found equal to 0.438 ± 0.009 μs and the effective delayed neutron fraction is 765 ± 8 pcm. Discrepancies with the calculated value (722 pcm, calculation from ENEA) are satisfactory: -5.6% for the Rossi-α estimate and -2.7% for the Feynman-α estimate. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Parametric Modeling and Fault Tolerant Control

Fault tolerant control is considered for a nonlinear aircraft model expressed as a linear parameter-varying system. By proper parameterization of foreseeable faults, the linear parameter-varying system can include fault effects as additional varying parameters. A recently developed technique in fault effect parameter estimation allows us to assume that estimates of the fault effect parameters are available on-line. Reconfigurability is calculated for this model with respect to the loss of control effectiveness to assess the potentiality of the model to tolerate such losses prior to control design. The control design is carried out by applying a polytopic method to the aircraft model. An error bound on fault effect parameter estimation is provided, within which the Lyapunov stability of the closed-loop system is robust. Our simulation results show that as long as the fault parameter estimates are sufficiently accurate, the polytopic controller can provide satisfactory fault-tolerance.

Wu, N. Eva↗

Data Assimilation of Terrestrial Water Storage Observations to Estimate Precipitation Fluxes: A Synthetic Experiment

The Gravity Recovery and Climate Experiment (GRACE) mission and its Follow-On (GRACE-FO) mission provide unprecedented observations of terrestrial water storage (TWS) dynamics at basin to continental scales. Established GRACE data assimilation techniques directly adjust the simulated water storage components to improve the estimation of groundwater, streamflow, and snow water equivalent. Such techniques artificially add/subtract water to/from prognostic variables, thus upsetting the simulated water balance. To overcome this limitation, we propose and test an alternative assimilation scheme in which precipitation fluxes are adjusted to achieve the desired changes in simulated TWS. Using a synthetic data assimilation experiment, we show that the scheme improves performance skill in precipitation estimates in general, but that it is more robust for snowfall than for rainfall, and it fails in certain regions with strong horizontal gradients in precipitation. The results demonstrate that assimilation of TWS observations can help correct (adjust) the model’s precipitation forcing and, in turn, enhance model estimates of TWS, snow mass, soil moisture, runoff, and evaporation. A key limitation of the approach is the assumption that all errors in TWS originate from errors in precipitation. Nevertheless, the proposed approach produces more consistent improvements in simulated runoff than the established GRACE data assimilation techniques.

Data Assimilation↗

Kalman Filter for Spinning Spacecraft Attitude Estimation

This paper presents a Kalman filter using a seven-component attitude state vector comprising the angular momentum components in an inertial reference frame, the angular momentum components in the body frame, and a rotation angle. The relatively slow variation of these parameters makes this parameterization advantageous for spinning spacecraft attitude estimation. The filter accounts for the constraint that the magnitude of the angular momentum vector is the same in the inertial and body frames by employing a reduced six-component error state. Four variants of the filter, defined by different choices for the reduced error state, are tested against a quaternion-based filter using simulated data for the THEMIS mission. Three of these variants choose three of the components of the error state to be the infinitesimal attitude error angles, facilitating the computation of measurement sensitivity matrices and causing the usual 3x3 attitude covariance matrix to be a submatrix of the 6x6 covariance of the error state. These variants differ in their choice for the other three components of the error state. The variant employing the infinitesimal attitude error angles and the angular momentum components in an inertial reference frame as the error state shows the best combination of robustness and efficiency in the simulations. Attitude estimation results using THEMIS flight data are also presented.

Markley, F. Landis↗

Kalman Filter for Spinning Spacecraft Attitude Estimation

This paper presents a Kalman filter using a seven-component attitude state vector comprising the angular momentum components in an inertial reference frame, the angular momentum components in the body frame, and a rotation angle. The relatively slow variation of these parameters makes this parameterization advantageous for spinning spacecraft attitude estimation. The filter accounts for the constraint that the magnitude of the angular momentum vector is the same in the inertial and body frames by employing a reduced six-component error state. Four variants of the filter, defined by different choices for the reduced error state, are tested against a quaternion-based filter using simulated data for the THEMIS mission. Three of these variants choose three of the components of the error state to be the infinitesimal attitude error angles, facilitating the computation of measurement sensitivity matrices and causing the usual 3x3 attitude covariance matrix to be a submatrix of the 6x6 covariance of the error state. These variants differ in their choice for the other three components of the error state. The variant employing the infinitesimal attitude error angles and the angular momentum components in an inertial reference frame as the error state shows the best combination of robustness and efficiency in the simulations. Attitude estimation results using THEMIS flight data are also presented.

Markley, F. Landis↗

Multiparameter Constraints on Empirical Infrasound Period-Yield Relations for Bolides and Implications for Planetary Defense

How effective are methods for estimating bolide energies from infrasound signal period-yield relationships? A single global period–energy relation can obscure significant variability introduced by parameters such as the atmospheric Doppler wind profile and the bolide’s energy deposition profile as a function of altitude. Bolide speed, entry angle, burst altitude, and multiepisode fragmentation may all play a role in defining the detected period of the shockwave. By leveraging bolide light-curve data from the Center for Near Earth Object Studies, we re-examined the period–energy relation as a function of these parameters. Through a bootstrap approach, we show that various event subsets can deviate from widely cited period–energy models and we identify which specific conditions most strongly reshape the period–energy scaling. The results define both the fidelity and reliability of period–energy relations when no additional data beyond the infrasound record is available and improve the outcome when supporting data from bolide trajectories and light curves are included. Ultimately, these findings expand the scope of earlier models, providing a nuanced and robust framework for infrasound-only yield estimation under a range of bolide scenarios.

Planetary science↗

DESI mock challenge: constructing DESI galaxy catalogues based on FastPM simulations

Together with larger spectroscopic surveys such as the Dark Energy Spectroscopic Instrument (DESI), the precision of large-scale structure studies and thus the constraints on the cosmological parameters are rapidly improving. Therefore, one must build realistic simulations and robust covariance matrices. We build galaxy catalogues by applying a halo occupation distribution (HOD) model upon the FastPM simulations, such that the resulting galaxy clustering reproduces high-resolution N-body simulations. While the resolution and halo finder are different from the reference simulations, we reproduce the reference galaxy two-point clustering measurements – monopole and quadrupole – to a precision required by the DESI Year 1 emission line galaxy sample down to non-linear scales, i.e. $k\lt 0.5\, h\, \mathrm{Mpc}^{-1}$ or $s\gt 10\, \mathrm{Mpc}\, h^{-1}$. Furthermore, we compute covariance matrices based on the resulting FastPM galaxy clustering – monopole and quadrupole. We study for the first time the effect of fitting on Fourier conjugate (e.g. power spectrum) on the covariance matrix of the Fourier counterpart (e.g. correlation function). We estimate the uncertainties of the two parameters of a simple clustering model and observe a maximum variation of 20 per cent for the different covariance matrices. Nevertheless, for most studied scales the scatter is between 2 and 10 per cent. Consequently, using the current pipeline we can precisely reproduce the clustering of N-body simulations and the resulting covariance matrices provide robust uncertainty estimations against HOD fitting scenarios. We expect our methodology will be useful for the coming DESI data analyses and their extension for other studies.

79 ASTRONOMY AND ASTROPHYSICS↗

Candidate gravitational microlensing events for future direct lens imaging

The mass of the lenses giving rise to Galactic microlensing events can be constrained by measuring the relative lens-source proper motion and lens flux. The flux of the lens can be separated from that of the source, companions to the source, and unrelated nearby stars with high-resolution images taken when the lens and source are spatially resolved. For typical ground-based adaptive optics (AO) or space-based observations, this requires either inordinately long time baselines or high relative proper motions. We provide a list of microlensing events toward the Galactic bulge with high relative lens-source proper motion that are therefore good candidates for constraining the lens mass with future high-resolution imaging. We investigate all events from 2004 to 2013 that display detectable finite-source effects, a feature that allows us to measure the proper motion. In total, we present 20 events with μ ≳ 8 mas yr{sup –1}. Of these, 14 were culled from previous analyses while 6 are new, including OGLE-2004-BLG-368, MOA-2005-BLG-36, OGLE-2012-BLG-0211, OGLE-2012-BLG-0456, MOA-2012-BLG-532, and MOA-2013-BLG-029. In ≲12 yr from the time of each event the lens and source of each event will be sufficiently separated for ground-based telescopes with AO systems or space telescopes to resolve each component and further characterize the lens system. Furthermore, for the most recent events, comparison of the lens flux estimates from images taken immediately to those estimated from images taken when the lens and source are resolved can be used to empirically check the robustness of the single-epoch method currently being used to estimate lens masses for many events.

79 ASTRONOMY AND ASTROPHYSICS↗

Robust time-domain identification of mass stiffness, and damping matrices

Accurate estimates of the mass, stiffness, and damping characteristics of a structure is necessary for determining the control laws best suited for active control methodologies. There are several modal identification techniques available for determining the frequencies, damping ratios, and mode shapes of a structure. However, modal identification methods in both the frequency and time domains have difficulties for certain circumstances. Frequency domain techniques which utilize the steady-state response from various harmonic inputs often encounter difficulties when the frequencies are closely distributed, the structure exhibits a high degree of damping, or the steady-state condition is hard to establish. Time domain techniques have produced successful results, but lack robustness with respect to measurement noise. In this paper, two identification techniques and an estimation method are combined to form a time-domain technique to accurately identify the mass, stiffness, and damping matrices from noisy measurements.

Roemer, Michael J.↗

Few-nucleon and many-nucleon systems with semilocal coordinate-space regularized chiral nucleon-nucleon forces

We employ a variety of ab initio methods, including Faddeev-Yakubovsky equations, no-core configuration interaction approach, coupled-cluster theory, and in-medium similarity renormalization group, to perform a comprehensive analysis of the nucleon-deuteron elastic and breakup reactions and selected properties of light and medium-mass nuclei up to 48 C a using the recently constructed semilocal coordinate-space regularized chiral nucleon-nucleon potentials. We compare the results with those based on selected phenomenological and chiral EFT two-nucleon potentials, discuss the convergence pattern of the chiral expansion, and estimate the achievable theoretical accuracy at various chiral orders using an approach to quantify truncation errors of the chiral expansion without relying on cutoff variation. We also address the robustness of this method and explore alternative ways to estimate the theoretical uncertainty from the truncation of the chiral expansion.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗