Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “error 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 397 records · Page 22

On the Training and Generalization of Deep Operator Networks

Here, we present a novel training method for deep operator networks (DeepONets), one of the most popular neural network models for operators. DeepONets are constructed by two subnetworks, namely the branch and trunk networks. Typically, the two subnetworks are trained simultaneously, which amounts to solving a complex optimization problem in a high dimensional space. In addition, the nonconvex and nonlinear nature makes training very challenging. To tackle such a challenge, we propose a two-step training method that trains the trunk network first and then sequentially trains the branch network. The core mechanism is motivated by the divide-and-conquer paradigm and is the decomposition of the entire complex training task into two subtasks with reduced complexity. Therein the Gram–Schmidt orthonormalization process is introduced which significantly improves stability and generalization ability. On the theoretical side, we establish a generalization error estimate in terms of the number of training data, the width of DeepONets, and the number of input and output sensors. Numerical examples are presented to demonstrate the effectiveness of the two-step training method, including Darcy flow in heterogeneous porous media.

deep operator networks↗

A new computational approach for the solutions of generalized pantograph-delay differential equations

In this study, a new computational approach is presented to solve the generalized pantograph-delay differential equations (PDDEs). The solutions obtained by our scheme represent by a linear combination of a special kind of basis functions, and can be deduced in a straightforward manner. Firstly, using the least squares approximation method and the Lagrange-multiplier method, the given PDDE is converted to a linear system of algebraic equations, and those unknown coefficients of the solution of the problem are determined by solving this linear system. Secondly, a PDDE related to the error function of the approximate solution is constructed based on the residual error function technique, and error estimation is presented for the suggested method. The convergence of the approximate solution is proved. Several numerical examples are given to demonstrate the accuracy and efficiency. Comparisons are made between the proposed method and other existing methods.

97 MATHEMATICS AND COMPUTING↗

A Fast Solver for the Fractional Helmholtz Equation

The purpose of this paper is to study a Helmholtz problem with a spectral fractional Laplacian, instead of the standard Laplacian. Recently, it has been established that such a fractional Helmholtz problem better captures the underlying behavior in Geophysical Electromagnetics. In this work, we establish the well-posedness and regularity of this problem. We introduce a hybrid spectral-finite element approach to discretize it and show well-posedness of the discrete system. In addition, we derive a priori discretization error estimates. Finally, we introduce an efficient solver that scales aswell as the best possible solver for the classical integer-order Helmholtz equation. We conclude withseveral illustrative examples that confirm our theoretical findings.

97 MATHEMATICS AND COMPUTING↗

Error and Correction Analysis for the FFA@CEBAF Energy Upgrade

An energy upgrade design for the Continuous Electron Beam Accelerator Facility (CEBAF) is under development, using fixed field alternating gradient (FFA) return arcs to recirculate electron beam up to an additional five times through the accelerating structures at CEBAF. A necessary component of any large accelerator is a beam steering and optical correction system. Small environmental changes and system errors can lower beam quality or even shut down the machine; and in pursuit of the scientific mission of JLab, high quality electron beams must be delivered to the experimental halls on a predictable schedule. Correction in the novel FFA arcs of the current upgrade design is complicated by several factors. These complexities inform the choice of correction algorithm structure and parameter values. A baseline algorithm in addition to diagnostic and correction hardware configuration is presented. The effect of this correction protocol is shown with respect to estimated errors, and several possible extensions of the algorithm are discussed. This work presents an important proof of concept for the FFA@CEBAF design effort, and provides a functional correction strategy which may be simply adjusted and optimized for future design changes.

Coxe, Alex [Old Dominion Univ., Norfolk, VA (Unite↗

Development of an optical three-dimensional laser tracker using dual modulated laser diodes and a signal detector

Laser trackers are widely used in industry for tasks such as the assembly of airplanes and automobiles, contour measurement, and robot calibration. However, laser trackers are expensive, and the corresponding solution procedure is very complex. The influence of measurement uncertainties is also significant. This study proposes a three-dimensional space position measurement system which consists of two tracking modules, a zero tracking angle return subsystem, and a target quadrant photodiode (QPD). The target QPD is placed on the object being tracked. The origin locking method is used to keep the rays on the origin of the target QPD. The position of the target QPD is determined using triangulation since the two laser rays are projected onto one QPD. Modulation and demodulation are utilized to separate the coupled positional values. The experiment results show that measurement errors in the X, Y, and Z directions are less than {+-}0.05% when the measured object was moved by 300, 300, and 200 mm in the X, Y, and Z axes, respectively. The theoretical measurement error estimated from the measurement model is between {+-}0.02% and {+-}0.07% within the defined measurable range. The proposed system can be applied to the measurements of machine tools and robot arms.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Primordial power spectrum from Planck

Using modified Richardson-Lucy algorithm we reconstruct the primordial power spectrum (PPS) from Planck Cosmic Microwave Background (CMB) temperature anisotropy data. In our analysis we use different combinations of angular power spectra from Planck to reconstruct the shape of the primordial power spectrum and locate possible features. Performing an extensive error analysis we found the dip near ℓ ∼ 750–850 represents the most prominent feature in the data. Feature near ℓ ∼ 1800–2000 is detectable with high confidence only in 217 GHz spectrum and is apparently consequence of a small systematic as described in the revised Planck 2013 papers. Fixing the background cosmological parameters and the foreground nuisance parameters to their best fit baseline values, we report that the best fit power law primordial power spectrum is consistent with the reconstructed form of the PPS at 2σ C.L. of the estimated errors (apart from the local features mentioned above). As a consistency test, we found the reconstructed primordial power spectrum from Planck temperature data can also substantially improve the fit to WMAP-9 angular power spectrum data (with respect to power-law form of the PPS) allowing an overall amplitude shift of ∼ 2.5%. In this context low-ℓ and 100 GHz spectrum from Planck which have proper overlap in the multipole range with WMAP data found to be completely consistent with WMAP-9 (allowing amplitude shift). As another important result of our analysis we do report the evidence of gravitational lensing through the reconstruction analysis. Finally we present two smooth form of the PPS containing only the important features. These smooth forms of PPS can provide significant improvements in fitting the data (with respect to the power law PPS) and can be helpful to give hints for inflationary model building.

79 ASTRONOMY AND ASTROPHYSICS↗

Final Report for Project "Framework Application for Core-Edge Transport Simulations (FACETS)"

This is the final report for the Colorado State University Component of the FACETS Project. FACETS was focused on the development of a multiphysics, parallel framework application that could provide the capability to enable whole-device fusion reactor modeling and, in the process, the development of the modeling infrastructure and computational understanding needed for ITER. It was intended that FACETS be highly flexible, through the use of modern computational methods, including component technology and object oriented design, to facilitate switching from one model to another for a given aspect of the physics, and making it possible to use simplified models for rapid turnaround or high-fidelity models that will take advantage of the largest supercomputer hardware. FACETS was designed in a heterogeneous parallel context, where different parts of the application can take advantage through parallelism based on task farming, domain decomposition, and/or pipelining as needed and applicable. As with all fusion simulations, an integral part of the FACETS project was treatment of the coupling of different physical processes at different scales interacting closely. A primary example for the FACETS project is the coupling of existing core and edge simulations, with the transport and wall interactions described by reduced models. However, core and edge simulations themselves involve significant coupling of different processes with large scale differences. Numerical treatment of coupling is impacted by a number of factors including, scale differences, form of information transferred between processes, implementation of solvers for different codes, and high performance computing concerns. Operator decomposition involving the computation of the individual processes individually using appropriate simulation codes and then linking/synchronizing the component simulations at regular points in space and time, is the defacto approach to high performance simulation of multiphysics, multiscale systems. Various forms of operator decomposition are used in nearly all fusion simulations. However, operator decomposition generally has a complex effect on accuracy and stability of numerical simulations. Yet, these effects can be difficult to detect. The Colorado State University component of the FACETS team led by P.I.D. Estep was focused on analyzing the effects of operator decomposition on fusion simulations. The approach was based on a posteriori error analysis employing adjoint problems, computable residuals, and variational analysis to produce accurate computational error estimates for quantities of interest. Computable residuals are used to quantify the effects of various discretization choices. The generalized Greens function satisfying the adjoint problem quantities the effects of stability. Technical issues to be addressed included: (1) defining appropriate adjoint operators for operator decomposition discretizations; (2) determining the appropriate residuals for the multifaceted aspects involved with multiphysics discretizations; (3) producing the estimates within the computational framework of existing fusion codes; (4) carrying out the analysis for discretizations used in fusion simulations; and (5) devising efficient approaches to mitigating the effects of discretization. This report provides a summary of the accomplished research and a detailed description of personnel, activities, outcomes and achievements.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOES THE MILKY WAY OBEY SPIRAL GALAXY SCALING RELATIONS?

It is crucial to understand how the Milky Way (MW), the galaxy we can study in the most intimate detail, fits in among other galaxies. Key considerations include the Tully–Fisher relation (TFR)—i.e., the tight correlation between luminosity ( L ) and rotational velocity ( V {sub rot})—and the three-dimensional luminosity–velocity–radius ( LVR ) scaling relation. Several past studies have characterized the MW as a 1–1.5 σ outlier to the TFR. This study re-examines such comparisons using new estimates of MW properties that are robust to many of the systematic uncertainties that have been a problem in the past and are based on assumptions consistent with those used for other spiral galaxies. Comparing to scaling relations derived from modern extragalactic data, we find that our Galaxy’s properties are in excellent agreement with TFRs defined using any Sloan Digital Sky Survey-filter absolute magnitude, stellar mass, or baryonic mass as the L proxy. We next utilize disk scale length ( R {sub d}) measurements to extend this investigation to the LVR relation. Here we find that our Galaxy lies farther from the relation than ∼90% of other spiral galaxies, yielding ∼9.5 σ evidence that it is unusually compact for its L and V {sub rot} (based on MW errors alone), a result that holds for all of the L proxies considered. The expected R {sub d} for the MW from the LVR relation is ∼5 kpc, nearly twice as large as the observed value, with error estimates placing the two in tension at the ∼1.4 σ level. The compact scale length of the Galactic disk could be related to other ways in which the MW has been found to be anomalous.

79 ASTRONOMY AND ASTROPHYSICS↗

PIROCK: A swiss-knife partitioned implicit–explicit orthogonal Runge–Kutta Chebyshev integrator for stiff diffusion–advection–reaction problems with or without noise

A partitioned implicit–explicit orthogonal Runge–Kutta method (PIROCK) is proposed for the time integration of diffusion–advection–reaction problems with possibly severely stiff reaction terms and stiff stochastic terms. The diffusion terms are solved by the explicit second order orthogonal Chebyshev method (ROCK2), while the stiff reaction terms (solved implicitly) and the advection and noise terms (solved explicitly) are integrated in the algorithm as finishing procedures. It is shown that the various coupling (between diffusion, reaction, advection and noise) can be stabilized in the PIROCK method. The method, implemented in a single black-box code that is fully adaptive, provides error estimators for the various terms present in the problem, and requires from the user solely the right-hand side of the differential equation. Numerical experiments and comparisons with existing Chebyshev methods, IMEX methods and partitioned methods show the efficiency and flexibility of our new algorithm.

97 MATHEMATICS AND COMPUTING↗

Milstein Approximation for Advection-Diffusion Equations Driven by Multiplicative Noncontinuous Martingale Noises

In this paper, the strong approximation of a stochastic partial differential equation, whose differential operator is of advection-diffusion type and which is driven by a multiplicative, infinite dimensional, cadlag, square integrable martingale, is presented. A finite dimensional projection of the infinite dimensional equation, for example a Galerkin projection, with nonequidistant time stepping is used. Error estimates for the discretized equation are derived in L{sup 2} and almost sure senses. Besides space and time discretizations, noise approximations are also provided, where the Milstein double stochastic integral is approximated in such a way that the overall complexity is not increased compared to an Euler-Maruyama approximation. Finally, simulations complete the paper.

97 MATHEMATICS AND COMPUTING↗

LAGRANGIAN RELAXATION SCHEMES FOR CALCULATING FORCE-FREE MAGNETIC FIELDS, AND THEIR LIMITATIONS

Force-free magnetic fields are important in many astrophysical settings. Determining the properties of such force-free fields-especially smoothness and stability properties-is crucial to understanding many key phenomena in astrophysical plasmas, for example, energy release processes that heat the plasma and lead to dynamic or explosive events. In the present work we discuss a serious limitation on the computation of force-free fields, within the context of a Lagrangian relaxation scheme that conserves magnetic flux and {nabla} {center_dot} B identically. This issue has the potential to invalidate the results produced by numerical force-free field solvers even for cases in which they appear to converge (at fixed grid resolution) to an equilibrium magnetic field. Error estimates are introduced to assess the quality of the calculated equilibrium. We go on to present an algorithm, based on rewriting the curl operation via Stokes' theorem, for calculating the current which holds great promise for improving dramatically the accuracy of the Lagrangian relaxation procedure.

99 GENERAL AND MISCELLANEOUS↗

Dynamic signal recovery in distribution grids using compressive lossy measurements

Distribution system state estimation requires reliable aggregation of the measured data. However, the large volume of the measured data imposes a significant stress on the underlying communication infrastructure. With the challenges associated with measurement availability, current distribution systems are typically unobservable. To cope with the unobservability issue, compressive sensing theory allows us to recover system state information from a small number of measurements provided the states of the distribution system exhibit sparsity. In this paper, we evaluate the robustness of an updated Kalman filtered modified compressive sensing (KF-ModCS) technique that dynamically estimates the grid states using a small fraction of measured data. In practice, measurements used for sparsity based state estimation may also be intermittent due to communication network induced losses. Further, to understand the effect of packet losses on KF-ModCS, we provide an upper bound for the expected variances of the state estimation error for a given rate of information loss. This upper bound is further improved if the support set of the sparse signal that characterizes the state dynamics does not change over time and/or the reduced model is observable. Simulations based on two practical data sets collected from actual customers in a distribution grid validate the theoretical results.

24 POWER TRANSMISSION AND DISTRIBUTION↗

HORIZON RUN 3: TOPOLOGY AS A STANDARD RULER

We study the physically self-bound cold dark matter halo distribution, which we associate with the massive galaxies within Horizon Run 3, to estimate the accuracy of the determination of the cosmological distance scale measured by the topology analysis. We apply the routine '''Contour 3D''' to the 108 Mock Survey of π steradians out to redshift z = 0.6, which effectively corresponds to the SDSS-III Baryon Oscillation Spectroscopic Survey (BOSS) survey, and compare the topology with that of a Gaussian random phase field. We find that given three separate smoothing lengths λ = 15, 21, and 34 h {sup –1} Mpc, the least χ{sup 2} fit genus per unit volume (g) yields a 1.7% fractional uncertainty in smoothing length and angular diameter distance to z = 0.6. This is an improvement on former calibrations and presents an error estimate competitive with baryon acoustic oscillation scale techniques. We also present three-dimensional graphics of the Horizon Run 3 spherical mock survey to show a wealth of large-scale structures of the universe that are expected for surveys like BOSS.

79 ASTRONOMY AND ASTROPHYSICS↗

Nonparametric reconstruction of the dark energy equation of state from diverse data sets

The cause of the accelerated expansion of the Universe poses one of the most fundamental questions in physics today. In the absence of a compelling theory to explain the observations, a first task is to develop a robust phenomenological approach: If the acceleration is driven by some form of dark energy, then the phenomenology is determined by the form of the dark energy equation of state w(z) as a function of redshift. A major aim of ongoing and upcoming cosmological surveys is to measure w and its evolution at high accuracy. Since w(z) is not directly accessible to measurement, powerful reconstruction methods are needed to extract it reliably from observations. We have recently introduced a new reconstruction method for w(z) based on Gaussian process modeling. This method can capture nontrivial w(z) dependences and, most importantly, it yields controlled and unbiased error estimates. In this paper we extend the method to include a diverse set of measurements: baryon acoustic oscillations, cosmic microwave background measurements, and supernova data. We analyze currently available data sets and present the resulting constraints on w(z), finding that current observations are in very good agreement with a cosmological constant. In addition, we explore how well our method captures nontrivial behavior of w(z) by analyzing simulated data assuming high-quality observations from future surveys. We find that the baryon acoustic oscillation measurements by themselves already lead to remarkably good reconstruction results and that the combination of different high-quality probes allows us to reconstruct w(z) very reliably with small error bounds.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A robust dynamic state estimation approach against model errors caused by load changes

Dynamic state estimation (DSE) plays an important role in power system security monitoring and online control. In practice, there are two approaches to implementing DSE. The first approach is distributed DSE, which is based on the assumption that the terminal bus of each generator can be measured by PMUs (phasor measurement units). The assumption cannot be satisfied currently, however, because PMUs usually are installed at important high-voltage buses such as 500-kV buses installed in portions of the grid overseen by the Western Electricity Coordinating Council. Another issue of this approach is that performance of DSE is vulnerable to bad measurement data. The reason for this vulnerability is that DSE is performed separately through measurements at each terminal bus, and measurements at terminal buses are the only measurement upon which DSE can rely. Therefore, important redundant measurements are not included in this approach. The second approach is centralized DSE. This approach does not have the requirement for PMU location, and redundant measurements can be considered fully. However, load changes and grid topology changes impact centralized DSE. In this paper, we propose a new approach for handling the impact of load changes on DSE. We have developed a new algorithm that includes two sequential steps. In the first step, errors caused by load changes are detected by analyzing the difference between prediction results and measured results. In the second step, once model error is detected, a model optimization procedure is run to correct the error so the state estimation error can be mitigated. Simulation results from the IEEE 68 bus system show that the proposed approach can effectively handle model errors caused by load changes.

robust dynamic state estimation, load change, powe↗

THE COLOR-PERIOD DIAGRAM AND STELLAR ROTATIONAL EVOLUTION-NEW ROTATION PERIOD MEASUREMENTS IN THE OPEN CLUSTER M34

We present the results of a 5 month photometric time-series survey for stellar rotation periods combined with a 4 year radial-velocity survey for membership and binarity in the 220 Myr open cluster M34. We report surface rotation periods for 120 stars, 83 of which are kinematic and photometric late-type cluster members. A comparison to previous work serves to illustrate the importance of high-cadence long baseline photometric observations and membership information. The new M34 periods are less biased against slow rotation and cleaned for non-members. The rotation periods of the cluster members span over more than an order of magnitude from 0.5 days up to 11.5 days, and trace two distinct rotational sequences-fast (C) and moderate-to-slow (I)-in the color-period diagram. The sequences represent two different states (fast and slow) in the rotational evolution of the late-type cluster members. We use the color-period diagrams for M34 and for younger and older clusters to estimate the timescale for the transition from the C to the I sequence and find {approx}<150 Myr, {approx}150-300 Myr, and {approx}300-600 Myr for G, early-mid K, and late K dwarfs, respectively. The small number of stars in the gap between C and I suggests a quick transition. We estimate a lower limit on the maximum spin-down rate (dP/dt) during this transition to be {approx}0.06 days Myr{sup -1} and {approx}0.08 days Myr{sup -1} for early and late K dwarfs, respectively. We compare the I sequence rotation periods in M34 and the Hyades for G and K dwarfs and find that K dwarfs spin down slower than the Skumanich {radical}t rate. We determine a gyrochronology age of 240 Myr for M34. The gyro-age has a small formal uncertainty of 2% which reflects the tight I sequence in the M34 color-period diagram. We measure the effect of cluster age uncertainties on the gyrochronology age for M34 and find the resulting error on the gyro-age to be consistent with the {approx}15% error estimate for the technique in general. We use the M34 I sequence to redetermine the coefficients in the expression for rotational dependence on color used in gyrochronology. Finally, we propose that stability in the phase, shape, and amplitude of the photometric variability for the 120 rotators over the {approx}5 month duration of our survey is due to spot generation at active stellar longitudes.

79 ASTRONOMY AND ASTROPHYSICS↗

Strict Convexity of the Free Energy of the Canonical Ensemble Under Decay of Correlations

We consider a one-dimensional lattice system of unbounded, real-valued spins. We allow arbitrary strong, attractive, nearest-neighbor interaction. We show that the free energy of the canonical ensemble (ce) converges uniformly in C{sup 2} to the free energy of the grand ce (gce). The error estimates are quantitative. A direct consequence is that the free energy of the ce is uniformly strictly convex for large systems. Another consequence is a quantitative local Cramér theorem which yields the strict convexity of the coarse-grained Hamiltonian. With small adaptations, the argument could be generalized to systems with finite-range interaction on a graph, as long as the degree of the graph is uniformly bounded and the associated gce has uniform decay of correlations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Assessing the inner core nucleation paradox with atomic-scale simulations

We investigate the conditions required to freeze liquid iron and iron alloys near the centre of Earth's core. It is usually assumed that inner core growth begins once the ambient core temperature falls below the melting temperature of the iron alloy at Earth's centre; however, additional (under)cooling is required to overcome the energy barrier associated with creating a solid–liquid interface. Predictions based on Classical Nucleation Theory (CNT) have estimated a required undercooling of ~1000 K, which cannot be reconciled with predicted core cooling rates of ~100 K Gyr -1 . This apparent contradiction has been called the ‘inner core nucleation paradox’. Here we address three major uncertainties in the application of CNT to inner core nucleation using atomic-scale simulations. First, we simulate freezing in Fe and Fe-O liquids at core conditions to self-consistently constrain all parameters required by the CNT equations. Second, we test the basic validity of CNT by directly calculating the waiting time to observe freezing events in Fe and Fe-O liquids. Third, we investigate the influence of wave-like forcings applied to the atomic simulations, which have been suggested as a means to significantly reduce the energy barrier. Our results are consistent with CNT in the computationally accessible parameter regime, though error estimates on the waiting time can reach 50% of the measurement at the largest undercooling temperatures. Using CNT to extrapolate to inner core conditions yields estimated undercooling of 730 ± 20 K for the pure iron system and 675 ± 35 K for the Fe-O system. Forcings corresponding to large pressure variations of O(10) GPa reduce these values by ~100 K. While our undercooling estimates are significantly lower than previous estimates they are not low enough to resolve the inner core nucleation paradox.

58 GEOSCIENCES↗