Search NASA⌕ Search

SEARCH · Search NASA

Results for “Tikhonov regularization”

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 55 records · Page 3

Smooth full field reconstruction of velocity and its gradients from noisy scattered velocimetry data in a cross-slot flow

We present a method combining generalized Tikhonov regularization with a finite element approximation for reconstructing smooth velocity and velocity gradient fields from spatially scattered and noisy velocity data in a two-dimensional complex flow domain. Synthetic velocity data for a cross-slot geometry are generated using the Oldroyd-B solution, subsequently perturbed by random noise. Performances of diverse finite element continuity-regularization criterion combinations are tested against noise-free data, while the optimum regularization parameter is determined using generalized cross-validation. The best performance is achieved for the velocity field and its gradients simultaneously by C{sup 2} continuous Hermite finite elements and minimization of a norm of the velocity's third derivative. The standard regularization criterion based on the second derivative is shown to lead to systematic distortions in boundary regions, allowing therefore a lower reduction in the statistical error. Furthermore, optical fields are calculated by applying a differential constitutive equation directly to the reconstructed flow kinematics; high quality velocity gradient fields are shown to be an essential prerequisite for their reliable prediction. Overall, the method is expedient to implement and does not require boundary conditions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Separation of continuous lines mutually overlapping and smoothed by the instrumental function

Direct and inverse spectroscopy problems concerning superposition of overlapping continuous spectral lines, as well as smoothing of the total spectrum by the instrumental function, are considered. The direct problem is formulated in two stages: initially, the total spectrum is formed by lines with a given intensity distribution, and, then, a smoothed by the spectrometer instrumental function and noisy spectrum is obtained. The inverse problem is also formulated in two stages: initially, the Fredholm integral equation of the first kind is solved by the Tikhonov regularization method (an ill-posed problem), and, then, the problem of reconstruction of separate line-component shapes from the total spectrum is solved, which is the problem of line separation (division). The individual line components are modeled by Gaussians and Lorentzians. Numerical illustrations are shown. Gaussian and dispersion (Lorentz) instrumental functions are considered.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Multigrid Preconditioning of Linear Systems for Interior Point Methods Applied to a Class of Box-constrained Optimal Control Problems

We construct and analyze multigrid preconditioners for discretizations of operators of the form ${\mathcal D}_{\lambda}+{\mathcal K}^*{\mathcal K}$, where $D_{\lambda}$ is the multiplication with a relatively smooth function $\lambda>0$ and ${\mathcal K}$ is a compact linear operator. These systems arise when applying interior point methods to the minimization problem $\min_{u} \frac{1}{2}(|\!|{\mathcal K} u-f|\!|^2 +\beta|\!|u|\!|^2)$ with box-constraints $\underline{u}\leqslant u\leqslant\overline{u}$ on the controls. The presented preconditioning technique is closely related to the one developed by Draganescu and Dupont [Math. Comp., 77 (2008), pp. 2001–2038] for the associated unconstrained problem and is intended for large-scale problems. As in that work, the quality of the resulting preconditioners is shown to increase as $h\downarrow 0$, but it decreases as the smoothness of $\lambda$ declines. We test this algorithm on a Tikhonov-regularized backward parabolic equation with box-constraints on the control and on a standard elliptic-constrained optimization problem. In both cases it is shown that the number of linear iterations per optimization step, as well as the total number of finest-scale matrix-vector multiplications, is decreasing with increasing resolution, thus showing the method to be potentially very efficient for truly large-scale problems.

97 MATHEMATICS AND COMPUTING↗

ForceFinder

SAND2025-11750O ForceFinder extends the Structural Dynamics Python Libraries (SDynPy) with comprehensive tools for inverse source estimation (ISE) tasks via frequency response function (FRF) matrix inversion. The software is designed for transfer path analysis and multiple-input/multiple-output (MIMO) vibration control problems. It allows users to estimate sources through various algorithms, from the basic Moore-Penrose pseudo-inverse to statistical learning methods such as Tikhonov regularization via an L-curve and elastic net regularization via an information criterion. ForceFinder uses an object-oriented framework, where all components of the ISE problem—such as FRFs, responses, and transformations—are stored in a "SourcePathReceiver" object. This software can be applied to any noise and vibration problem. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Carter, Steven [Sandia National Lab. (SNL-CA), Liv↗

Many-Group Cross-Section Adjustment Techniques for Boiling Water Reactor Adaptive Simulation

Computational capability has been developed to adjust multigroup neutron cross sections, including self-shielding correction factors, to improve the fidelity of boiling water reactor (BWR) core modeling and simulation. The method involves propagating multigroup neutron cross-section uncertainties through various BWR computational models to evaluate uncertainties in key core attributes such as core k{sub eff}, nodal power distributions, thermal margins, and in-core detector readings. Uncertainty-based inverse theory methods are then employed to adjust multigroup cross sections to minimize the disagreement between BWR core modeling predictions and observed (i.e., measured) plant data. For this paper, observed plant data are virtually simulated in the form of perturbed three-dimensional nodal power distributions with the perturbations sized to represent actual discrepancies between predictions and real plant data. The major focus of this work is to efficiently propagate multigroup neutron cross-section uncertainty through BWR lattice physics and core simulator calculations. The data adjustment equations are developed using a subspace approach that exploits the ill-conditioning of the multigroup cross-section covariance matrix to minimize computation and storage burden. Tikhonov regularization is also employed to improve the conditioning of the data adjustment equations. Expressions are also provided for posterior covariance matrices of both the multigroup cross-section and core attributes uncertainties.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

PyRAT (Python Radiography Analysis Tool): Overview [Slides]

PyRAT was developed as a quantitative tool for robustly characterizing objects from radiographs to solve problems such as the hybrid nonlinear inverse problem. The optimization software library that was used is the nonsmooth optimization by MADS algorithm (NOMAD). Some of PyRAT's features are: (1) hybrid nonlinear inverse problem with calculated x-ray spectrum and detector response; (2) optimization based inversion approach with goal of identifying unknown object configurations - MVO problem; (3) using functionalities of Python libraries for radiographic image processing and analysis; (4) using the Tikhonov regularization method of linear inverse problem to recover partial information of object configurations; (5) using a priori knowledge of problem solutions to define feasible region and discrete neighbor for the MVO problem - initial data analysis + material library $\Rightarrow$ a priori knowledge; and (6) using the NOMAD (C++ version) software in the object.

99 GENERAL AND MISCELLANEOUS↗

Advanced Methods for Modeling Water-Levels and Estimating Drawdowns with SeriesSEE, an Excel Add-In

Water-level modeling is used for multiple-well aquifer tests to reliably differentiate pumping responses from natural water-level changes in wells, or environmental fluctuations. Synthetic water levels are created during water-level modeling and represent the summation of multiple component fluctuations, including those caused by environmental forcing and pumping. Pumping signals are modeled by transforming step-wise pumping records into water-level changes by using superimposed Theis functions. Water-levels can be modeled robustly with this Theis-transform approach because environmental fluctuations and pumping signals are simulated simultaneously. Water-level modeling with Theis transforms has been implemented in the program SeriesSEE, which is a Microsoft® Excel add-in. Moving average, Theis, pneumatic-lag, and gamma functions transform time series of measured values into water-level model components in SeriesSEE. Earth tides and step transforms are additional computed water-level model components. Water-level models are calibrated by minimizing a sum-of-squares objective function where singular value decomposition and Tikhonov regularization stabilize results. Drawdown estimates from a water-level model are the summation of all Theis transforms minus residual differences between synthetic and measured water levels. The accuracy of drawdown estimates is limited primarily by noise in the data sets, not the Theis-transform approach. Drawdowns much smaller than environmental fluctuations have been detected across major fault structures, at distances of more than 1 mile from the pumping well, and with limited pre-pumping and recovery data at sites across the United States. In addition to water-level modeling, utilities exist in SeriesSEE for viewing, cleaning, manipulating, and analyzing time-series data.

58 GEOSCIENCES↗

Interpretation of Flow Logs from Nevada Test Site Boreholes to Estimate Hydraulic conductivity Using Numerical Simulations Constrained by Single-Well Aquifer Tests

Hydraulic conductivities of volcanic and carbonate lithologic units at the Nevada Test Site were estimated from flow logs and aquifer-test data. Borehole flow and drawdown were integrated and interpreted using a radial, axisymmetric flow model, AnalyzeHOLE. This integrated approach is used because complex well completions and heterogeneous aquifers and confining units produce vertical flow in the annular space and aquifers adjacent to the wellbore. AnalyzeHOLE simulates vertical flow, in addition to horizontal flow, which accounts for converging flow toward screen ends and diverging flow toward transmissive intervals. Simulated aquifers and confining units uniformly are subdivided by depth into intervals in which the hydraulic conductivity is estimated with the Parameter ESTimation (PEST) software. Between 50 and 150 hydraulic-conductivity parameters were estimated by minimizing weighted differences between simulated and measured flow and drawdown. Transmissivity estimates from single-well or multiple-well aquifer tests were used to constrain estimates of hydraulic conductivity. The distribution of hydraulic conductivity within each lithology had a minimum variance because estimates were constrained with Tikhonov regularization. AnalyzeHOLE simulated hydraulic-conductivity estimates for lithologic units across screened and cased intervals are as much as 100 times less than those estimated using proportional flow-log analyses applied across screened intervals only. Smaller estimates of hydraulic conductivity for individual lithologic units are simulated because sections of the unit behind cased intervals of the wellbore are not assumed to be impermeable, and therefore, can contribute flow to the wellbore. Simulated hydraulic-conductivity estimates vary by more than three orders of magnitude across a lithologic unit, indicating a high degree of heterogeneity in volcanic and carbonate-rock units. The higher water transmitting potential of carbonate-rock units relative to volcanic-rock units is exemplified by the large difference in their estimated maximum hydraulic conductivity; 4,000 and 400 feet per day, respectively. Simulated minimum estimates of hydraulic conductivity are inexact and represent the lower detection limit of the method. Minimum thicknesses of lithologic intervals also were defined for comparing AnalyzeHOLE results to hydraulic properties in regional ground-water flow models.

58 GEOSCIENCES↗

Inversion of geothermal heat flux in a thermomechanically coupled nonlinear Stokes ice sheet model

We address the inverse problem of inferring the basal geothermal heat flux from surface velocity observations using a steady-state thermomechanically coupled nonlinear Stokes ice flow model. This is a challenging inverse problem since the map from basal heat flux to surface velocity observables is indirect: the heat flux is a boundary condition for the thermal advection–diffusion equation, which couples to the nonlinear Stokes ice flow equations; together they determine the surface ice flow velocity. This multiphysics inverse problem is formulated as a nonlinear least-squares optimization problem with a cost functional that includes the data misfit between surface velocity observations and model predictions. A Tikhonov regularization term is added to render the problem well posed. We derive adjoint-based gradient and Hessian expressions for the resulting partial differential equation (PDE)-constrained optimization problem and propose an inexact Newton method for its solution. As a consequence of the Petrov–Galerkin discretization of the energy equation, we show that discretization and differentiation do not commute; that is, the order in which we discretize the cost functional and differentiate it affects the correctness of the gradient. Using two- and three-dimensional model problems, we study the prospects for and limitations of the inference of the geothermal heat flux field from surface velocity observations. The results show that the reconstruction improves as the noise level in the observations decreases and that short-wavelength variations in the geothermal heat flux are difficult to recover. We analyze the ill-posedness of the inverse problem as a function of the number of observations by examining the spectrum of the Hessian of the cost functional. Motivated by the popularity of operator-split or staggered solvers for forward multiphysics problems – i.e., those that drop two-way coupling terms to yield a one-way coupled forward Jacobian – we study the effect on the inversion of a one-way coupling of the adjoint energy and Stokes equations. Here, we show that taking such a one-way coupled approach for the adjoint equations can lead to an incorrect gradient and premature termination of optimization iterations. This is due to loss of a descent direction stemming from inconsistency of the gradient with the contours of the cost functional. Nevertheless, one may still obtain a reasonable approximate inverse solution particularly if important features of the reconstructed solution emerge early in optimization iterations, before the premature termination.

15 GEOTHERMAL ENERGY↗

On compactness of admissible parameter sets: Convergence and stability in inverse problems for distributed parameter systems

A series of numerical examples is reported and several algorithms compared for estimation of coefficients in differential equation models. Unconstrained, constrained and Tikhonov regularization methods are tested for their behavior with regard to both convergence (of approximation methods for the states and parameters) and stability (continuity of the estimates with respect to perturbations in the data or observed states).

Banks, H. T.↗

Numerical recovery of material parameters in Euler-Bernoulli beam models

A fully Sinc-Galerkin method for recovering the spatially varying stiffness parameter in fourth-order time-dependence problems with fixed and cantilever boundary conditions is presented. The forward problems are discretized with a sinc basis in both the spatial and temporal domains. This yields an approximation solution which converges exponentially and is valid on the infinite time interval. When the forward methods are applied to parameter recovery problems, the resulting inverse problems are ill-posed. Tikhonov regularization is applied and the resulting minimization problems are solved via a quasi-Newton/trust region algorithm. The L-curve method is used to determine an appropriate value of the regularization parameter. Numerical results which highlight the method are given for problems with both fixed and cantilever boundary conditions.

Smith, R. C.↗

A fully Galerkin method for the recovery of stiffness and damping parameters in Euler-Bernoulli beam models

A fully Sinc-Galerkin method for recovering the spatially varying stiffness and damping parameters in Euler-Bernoulli beam models is presented. The forward problems are discretized with a sinc basis in both the spatial and temporal domains thus yielding an approximate solution which converges exponentially and is valid on the infinite time interval. Hence the method avoids the time-stepping which is characteristic of many of the forward schemes which are used in parameter recovery algorithms. Tikhonov regularization is used to stabilize the resulting inverse problem, and the L-curve method for determining an appropriate value of the regularization parameter is briefly discussed. Numerical examples are given which demonstrate the applicability of the method for both individual and simultaneous recovery of the material parameters.

Smith, R. C.↗

Sinc-Galerkin estimation of diffusivity in parabolic problems

A fully Sinc-Galerkin method for the numerical recovery of spatially varying diffusion coefficients in linear partial differential equations is presented. Because the parameter recovery problems are inherently ill-posed, an output error criterion in conjunction with Tikhonov regularization is used to formulate them as infinite-dimensional minimization problems. The forward problems are discretized with a sinc basis in both the spatial and temporal domains thus yielding an approximate solution which displays an exponential convergence rate and is valid on the infinite time interval. The minimization problems are then solved via a quasi-Newton/trust region algorithm. The L-curve technique for determining an approximate value of the regularization parameter is briefly discussed, and numerical examples are given which show the applicability of the method both for problems with noise-free data as well as for those whose data contains white noise.

Smith, Ralph C.↗

Coseismic fault slip associated with the 1992 M(sub w) 6.1 Joshua Tree, California, earthquake: Implications for the Joshua Tree-Landers earthquake sequence

Coseismic surface deformation associated with the M(sub w) 6.1, April 23, 1992, Joshua Tree earthquake is well represented by estimates of geodetic monument displacements at 20 locations independently derived from Global Positioning System and trilateration measurements. The rms signal to noise ratio for these inferred displacements is 1.8 with near-fault displacement estimates exceeding 40 mm. In order to determine the long-wavelength distribution of slip over the plane of rupture, a Tikhonov regularization operator is applied to these estimates which minimizes stress variability subject to purely right-lateral slip and zero surface slip constraints. The resulting slip distribution yields a geodetic moment estimate of 1.7 x 10(exp 18) N m with corresponding maximum slip around 0.8 m and compares well with independent and complementary information including seismic moment and source time function estimates and main shock and aftershock locations. From empirical Green's functions analyses, a rupture duration of 5 s is obtained which implies a rupture radius of 6-8 km. Most of the inferred slip lies to the north of the hypocenter, consistent with northward rupture propagation. Stress drop estimates are in the range of 2-4 MPa. In addition, predicted Coulomb stress increases correlate remarkably well with the distribution of aftershock hypocenters; most of the aftershocks occur in areas for which the mainshock rupture produced stress increases larger than about 0.1 MPa. In contrast, predicted stress changes are near zero at the hypocenter of the M(sub w) 7.3, June 28, 1992, Landers earthquake which nucleated about 20 km beyond the northernmost edge of the Joshua Tree rupture. Based on aftershock migrations and the predicted static stress field, we speculate that redistribution of Joshua Tree-induced stress perturbations played a role in the spatio-temporal development of the earth sequence culminating in the Landers event.

Bennett, Richard A.↗

The Detection of Radiated Modes from Ducted Fan Engines

The bypass duct of an aircraft engine is a low-pass filter allowing some spinning modes to radiate outside the duct. The knowledge of the radiated modes can help in noise reduction, as well as the diagnosis of noise generation mechanisms inside the duct. We propose a nonintrusive technique using a circular microphone array outside the engine measuring the complex noise spectrum on an arc of a circle. The array is placed at various axial distances from the inlet or the exhaust of the engine. Using a model of noise radiation from the duct, an overdetermined system of linear equations is constructed for the complex amplitudes of the radial modes for a fixed circumferential mode. This system of linear equations is generally singular, indicating that the problem is illposed. Tikhonov regularization is employed to solve this system of equations for the unknown amplitudes of the radiated modes. An application of our mode detection technique using measured acoustic data from a circular microphone array is presented. We show that this technique can reliably detect radiated modes with the possible exception of modes very close to cut-off.

Farassat, F.↗

Computation of the Streamfunction and Velocity Potential for Limited and Irregular Domains

An algorithm is proposed for the computation of streamfunction and velocity potential from given horizontal velocity vectors based on solving a minimization problem. To guarantee the uniqueness of the solution and computational reliability of the algorithm, a Tikhonov regularization is applied. The solution implies that the obtained streamfunction and velocity potential have minimal magnitude, while the given velocity vectors can be accurately reconstructed from the computed streamfunction and velocity potential. Because the formulation of the minimization problem allows for circumventing the explicit specification of separate boundary conditions on the streamfunction and velocity potential, the algorithm is easily applicable to irregular domains. By using an advanced minimization algorithm with the use of adjoint techniques, the method is computationally efficient and suitable for problems with large dimensions. An example is presented for coastal oceans to illustrate the practical application of the algorithm.

velocity↗

Extinction and Backscatter retrievals for High Spectral Resolution Lidar for high noise situations

High Spectral Resolution Lidars are uniquely valuable for characterizing atmospheric aerosol and clouds, both alone and in combination with passive instruments such as polarimeters. The benefits of HSRL are the ability to characterize the vertical variability of the atmosphere, to a greater degree than any passive instruments, and – compared to other lidar—additional channels of measurements that permit the direct retrieval of particulate extinction. HSRLs have been successfully deployed from both ground-based and airborne platforms, and will soon be included on satellites as well. From space, new challenges await, due to the need for fully au-tonomous cost-effective instrumentation and lower signal-to-noise ratios. Accordingly, algorithms for the retrieval of HSRL backscatter and extinction must also be reexamined and optimized for best performance in these more challenging circumstances. Motivated by the need for algorithms for the Clio lidar on the future Atmosphere Observing System polar platform, we explore and compare three methods of aerosol extinction retrieval for HSRL data: the direct differentiation method used in standard processing of LaRC airborne HSRL, optimal estimation, and Twomey-Tikhonov regularization. Pros and cons of each methodology will be dis-cussed and explored.

Sharon P. Burton↗

Decade-Long Ozone Profile Record from Suomi NPP OMPS Limb Profiler: Assessment of Version 2.6 Data

We evaluate a decadal ozone profile record derived from the Suomi National Polar-orbiting Partnership (SNPP) Ozone Mapping and Profiler Suite (OMPS) Limb Profiler (LP) satellite instrument. In 2023, the OMPS LP data were re-processed with the new version 2.6 retrieval algorithm that combines measurements from the ultraviolet (UV) and visible (VIS) parts of the spectra and employs the second order Tikhonov regularization to retrieve a single vertical ozone profile between 12.5 km (or cloud tops) and 57.5 km with the vertical resolution of about 1.9 - 2.5 km between 20-55 km. The algorithm uses radiances measured at six UV ozone-sensitive wavelengths (295, 302, 306, 312, 317 and 322 nm) paired with 353 nm, and one VIS wavelength at 606 nm combined with 510 nm and 675 nm to form a triplet. Each wavelength pair or triplet is used over a limited range of tangent altitudes where the sensitivity to ozone changes are strongest. A new implemented aerosol correction scheme is based on a gamma-function particle size distribution. Numerous calibration changes that affected ozone retrievals were also applied to measured LP radiances, including updates in altitude registration, radiometric calibration, stray light, and spectral registration. The key version 2.6 improvement is the reduction in relative drifts between LP ozone and correlative measurements, linked previously to a drift in the version 2.5 LP altitude registration. We compare LP ozone profiles with those from Aura Microwave Limb Sensor (MLS) to quantify ozone changes in version 2.6.

Limb scattering↗