Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “least square approximations”

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 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Interface PINNs (I-PINNs): A physics-informed neural networks framework for interface problems

Here, we present a novel physics-informed neural networks (PINNs) framework for modeling interface problems, termed Interface PINNs (I-PINNs). I-PINNs uses different neural networks for any two subdomains separated by a sharp interface such that the neural networks differ only through their activation functions while the other parameters remain identical. The performance of I-PINNs, conventional PINNs, and other existing domain-decomposition PINNs methods such as extended PINNs (XPINNs) and multi-domain PINN (M-PINN) is compared through several one-dimensional, two-dimensional, and three-dimensional benchmark elliptic interface problems. The results demonstrate that I-PINNs provides a root-mean-square-error accuracy, at least two orders of magnitude better than conventional PINNs and XPINNs at approximately one-tenth of the computational cost of conventional PINNs and half the cost of XPINNs. Additionally, while I-PINNs and M-PINN provide comparable accuracies, M-PINN is found to be approximately 50% more expensive.

42 ENGINEERING↗
DOE OSTI2024

A Kaczmarz-inspired approach to accelerate the optimization of neural network wavefunctions

Neural network wavefunctions optimized using the variational Monte Carlo method have been shown to produce highly accurate results for the electronic structure of atoms and small molecules, but the high cost of optimizing such wavefunctions prevents their application to larger systems. We propose the Subsampled Projected-Increment Natural Gradient Descent (SPRING) optimizer to reduce this bottleneck. SPRING combines ideas from the recently introduced minimum-step stochastic reconfiguration optimizer (MinSR) and the classical randomized Kaczmarz method for solving linear least-squares problems. We demonstrate that SPRING outperforms both MinSR and the popular Kronecker-Factored Approximate Curvature method (KFAC) across a number of small atoms and molecules, given that the learning rates of all methods are optimally tuned. For example, on the oxygen atom, SPRING attains chemical accuracy after forty thousand training iterations, whereas both MinSR and KFAC fail to do so even after one hundred thousand iterations.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1993

Identification of integrated airframe-propulsion effects on an F-15 aircraft for application to drag minimization

The application of an adaptive real-time measurement-based performance optimization technique is being explored for a future flight research program. The key technical challenge of the approach is parameter identification, which uses a perturbation-search technique to identify changes in performance caused by forced oscillations of the controls. The controls on the NASA F-15 highly integrated digital electronic control (HIDEC) aircraft were perturbed using inlet cowl rotation steps at various subsonic and supersonic flight conditions to determine the effect on aircraft performance. The feasibility of the perturbation-search technique for identifying integrated airframe-propulsion system performance effects was successfully shown through flight experiments and postflight data analysis. Aircraft response and control data were analyzed postflight to identify gradients and to determine the minimum drag point. Changes in longitudinal acceleration as small as 0.004 g were measured, and absolute resolution was estimated to be 0.002 g or approximately 50 lbf of drag. Two techniques for identifying performance gradients were compared: a least-squares estimation algorithm and a modified maximum likelihood estimator algorithm. A complementary filter algorithm was used with the least squares estimator.

Schkolnik, Gerald S.↗
NASA NTRS1993

Identification of integrated airframe: Propulsion effects on an F-15 aircraft for application to drag minimization

The application of an adaptive real-time measurement-based performance optimization technique is being explored for a future flight research program. The key technical challenge of the approach is parameter identification, which uses a perturbation-search technique to identify changes in performance caused by forced oscillations of the controls. The controls on the NASA F-15 highly integrated digital electronic control (HIDEC) aircraft were perturbed using inlet cowl rotation steps at various subsonic and supersonic flight conditions to determine the effect on aircraft performance. The feasibility of the perturbation-search technique for identifying integrated airframe-propulsion system performance effects was successfully shown through flight experiments and postflight data analysis. Aircraft response and control data were analyzed postflight to identify gradients and to determine the minimum drag point. Changes in longitudinal acceleration as small as 0.004 g were measured, and absolute resolution was estimated to be 0.002 g or approximately 50 lbf of drag. Two techniques for identifying performance gradients were compared: a least-squares estimation algorithm and a modified maximum likelihood estimator algorithm. A complementary filter algorithm was used with the least squares estimator.

Schkolnik, Gerard S.↗
NASA NTRS2007

Experimental Air-Broadened Line Parameters in the nu2 Band of CH3D

In this study we report the first experimental measurements of air-broadening and air-induced pressure-shift coefficients for approximately 378 transitions in the nu2 fundamental band of CH3D. These results were obtained from analysis of 17 room temperature laboratory absorption spectra recorded at 0.0056 cm(exp -1) resolution using the McMath-Pierce Fourier transform spectrometer located on Kitt Peak, Arizona. Three absorption cells with path lengths of 10.2, 25 and 150 cm were used to record the spectra. The total sample pressures ranged from 0.129x10(exp -2) to 52.855x10(exp -2) atm with CH3D volume mixing ratios of approximately 0.0109 in air. The spectra were analyzed using a multispectrum non-linear least-squares fitting technique. We report measurements for air pressure-broadening coefficients for transitions with quantum numbers as high as J" = 20 and K = 15, where K" = K' equivalent to K (for a parallel band). The measured air broadening coefficients range from 0.0205 to 0.0835 cm(exp -1) atm(exp -1) at 296 K. All the measured pressure-shift coefficients are negative and are found to vary from about -0.0005 to -0.0080 cm(exp -1) atm(exp -1) at the temperature of the spectra. We have examined the dependence of the measured broadening and shift parameters on the J" and K quantum numbers and also developed empirical expressions to describe the broadening coefficients in terms of m (m = -J", J" and J" + 1 in the (sup Q)P- (sup Q)Q-, and (sup Q)R-branch, respectively) and K. On average, the empirical expressions reproduce the measured broadening coefficients to within 4.4%.

Cross, Adriana Predoi↗
NASA NTRS1990

Global configuration of a magnetic cloud

A magnetic cloud associated with a 2N flare on January 1, 1978 was observed by IMP-8, Helios A, Helios B, and Voyager 2. The variation of the magnetic field observed at each spacecraft is represented to good approximation by Lundquist's solution for a cylindrically symmetric force-free magnetic field with constant alpha. A least-squares fit of Lundquist's solution to the data from each spacecraft gives the local orientation of the axis of the magnetic cloud. The times of the estimated boundaries of the magnetic cloud at each spacecraft, together with the speeds of the boundaries and the spacecraft position, give the positions of the boundaries at a given time. From these results the magnetic cloud is determined to resemble a flux rope whose minor radius is approximately 0.15 AU at 1 AU, and whose radius of curvature at 1 AU is approximately 1/3 AU.

Burlaga, L. F.↗
NASA NTRS1994

A simple suboptimal least-squares algorithm for attitude determination with multiple sensors

Three-axis attitude determination is equivalent to finding a coordinate transformation matrix which transforms a set of reference vectors fixed in inertial space to a set of measurement vectors fixed in the spacecraft. The attitude determination problem can be expressed as a constrained optimization problem. The constraint is that a coordinate transformation matrix must be proper, real, and orthogonal. A transformation matrix can be thought of as optimal in the least-squares sense if it maps the measurement vectors to the reference vectors with minimal 2-norm errors and meets the above constraint. This constrained optimization problem is known as Wahba's problem. Several algorithms which solve Wahba's problem exactly have been developed and used. These algorithms, while steadily improving, are all rather complicated. Furthermore, they involve such numerically unstable or sensitive operations as matrix determinant, matrix adjoint, and Newton-Raphson iterations. This paper describes an algorithm which minimizes Wahba's loss function, but without the constraint. When the constraint is ignored, the problem can be solved by a straightforward, numerically stable least-squares algorithm such as QR decomposition. Even though the algorithm does not explicitly take the constraint into account, it still yields a nearly orthogonal matrix for most practical cases; orthogonality only becomes corrupted when the sensor measurements are very noisy, on the same order of magnitude as the attitude rotations. The algorithm can be simplified if the attitude rotations are small enough so that the approximation sin(theta) approximately equals theta holds. We then compare the computational requirements for several well-known algorithms. For the general large-angle case, the QR least-squares algorithm is competitive with all other know algorithms and faster than most. If attitude rotations are small, the least-squares algorithm can be modified to run faster, and this modified algorithm is faster than all but a similarly specialized version of the QUEST algorithm. We also introduce a novel measurement averaging technique which reduces the n-measurement case to the two measurement case for our particular application, a star tracker and earth sensor mounted on an earth-pointed geosynchronous communications satellite. Using this technique, many n-measurement problems reduce to less than or equal to 3 measurements; this reduces the amount of required calculation without significant degradation in accuracy. Finally, we present the results of some tests which compare the least-squares algorithm with the QUEST and FOAM algorithms in the two-measurement case. For our example case, all three algorithms performed with similar accuracy.

Brozenec, Thomas F.↗
DOE OSTI2025

Kernel Manifolds: Nonlinear‐Augmentation Dimensionality Reduction Using Reproducing Kernel Hilbert Spaces

This paper generalizes recent advances on quadratic manifold (QM) dimensionality reduction by developing kernel methods-based nonlinear-augmentation dimensionality reduction. QMs, and more generally feature map-based nonlinear corrections, augment linear dimensionality reduction with a nonlinear correction term in the reconstruction map to overcome approximation accuracy limitations of purely linear approaches. While feature map-based approaches typically learn a least squares optimal polynomial correction term, we generalize this approach by learning an optimal nonlinear correction from a user-defined reproducing kernel Hilbert space. Our approach allows one to impose arbitrary nonlinear structure on the correction term, including polynomial structure, and includes feature map and radial basis function-based corrections as special cases. Furthermore, our method has relatively low training cost and has monotonically decreasing error as the latent space dimension increases. In conclusion, we compare our approach to proper orthogonal decomposition and several recent QM approaches on data from several example problems.

kernel methods↗
DOE OSTI2023

A Fourth-Order Embedded Boundary Finite Volume Method for the Unsteady Stokes Equations with Complex Geometries

A fourth-order finite volume embedded boundary (EB) method is presented for the unsteady Stokes equations. The algorithm represents complex geometries on a Cartesian grid using EB, employing a technique to mitigate the ``small cut-cell"" problem without mesh modifications, cell merging, or state redistribution. Spatial discretizations are based on a weighted least-squares technique that has been extended to fourth-order operators and boundary conditions, including an approximate projection to enforce the divergence-free constraint. Solutions are advanced in time using a fourth-order additive implicit-explicit Runge-Kutta method, with the viscous and source terms treated implicitly and explicitly, respectively. Formal accuracy of the method is demonstrated with several grid convergence studies, and results are shown for an application with a complex bio-inspired material. In conclusion, the developed method achieves fourth-order accuracy and is stable despite the pervasive small cells arising from complex geometries.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1977

Intensity of primary cosmic-ray electrons of energy exceeding 8 GeV

Results are reported for measurement of the intensity and energy spectrum of primary cosmic-ray electrons with a spark-chamber-counter-emulsion detector flown at a mean altitude of 3 g/sq cm residual atmosphere. A least-squares fit to the flight data yields an electron spectrum from 8 to 80 GeV of approximately 93E to the -2.91 power electrons/sq m/sec per sr/GeV. The results are compared with those of previous experiments as well as with the spectrum obtained for galactic nonthermal radiation. It is concluded that a 'clumpy' magnetic field proportional to the square root of matter density is consistent with measurements of high-energy electrons and synchrotron radiation toward the center of the Galaxy, that a gradual steepening of the electron spectrum relative to the proton spectrum is consistent with an electron lifetime of 1 million years, and that the density of cosmic-ray nucleons and electrons should be essentially uniform throughout the Galaxy if the nucleons have the same lifetime as the electrons and if they traversed 4 to 5 g/sq cm in that lifetime.

Freier, P.↗
NASA NTRS1984

Definitive orbit determination for the HEAO-2 spacecraft

Precise ephemerides for the High Energy Astronomy Observatory-2 (HEAO-2) were computed to assist in the Charles Stark Draper Laboratory development of an onboard orbit determination technique. Weighted least-squares, batch orbit solutions were calculated using a high-precision earth gravity model and an approximate model for intermittent spacecraft thrusting. With these improvements, orbit solution consistencies at the 50- to 100-meter level were attained using the Goddard Trajectory Determination System and NASA S-band tracking data.

Smith, R. L.↗
NASA NTRS1999

ASCA and Radio/RXTE Observations of GX 339-4

We have analyzed three separate archival Advanced Satellite for Cosmology and Astrophysics (ASCA) observations and eight separate Rossi X-ray Timing Explorer (RXTE) observations of the black hole candidate GX 339-4 in its low luminosity, spectrally hard state. Three of the RXTE observations were strictly simultaneous with 843 MHz and 8.3- 9.1 GHz radio observations. All data sets show evidence for an approximately 6.4 keV Fe line with equivalent widths approximately 20-100 eV. 'Reflection models' show a hardening of the RXTE spectra with decreasing X-ray flux; however, these models do not exhibit evidence of a correlation between the photon index of the incident power law flux and the solid angle subtended by the reflector. None of the models fit to the X-ray data, however, simultaneously explain the observed radio properties. We argue that the spatial extent of the observed radio emission is at least 0(10(exp 7 GM/c squared). Timing analysis reveals that all observations save one show evidence of a persistent f(qpo approximately equals 0.3 Hz quasi-periodic oscillations(quasi-periodic oscillations)). The broad band (10-3-102 Hz) power appears to be dominated by two independent processes that can be modeled as very broad Lorentzians with Q approximately less than 1. Similar to Cyg X-1, the hard photon variability is seen to lag the soft photon vaxiability with the lag time increasing with decreasing Fourier frequency. The magnitude of this time lag is seen to be positively correlated with the flux of GX 339-4.

Nowak, M. A.↗
NASA NTRS1993

Development of a gridless CFD method

The objective of this paper is to stimulate ideas concerning alternative approaches to discretizing fluid flow problems, i.e., the so-called gridless approach. The gridless approach uses only clouds of points and does not require that the points be connected to form a grid as is necessary in conventional CFD algorithms. The governing equations are solved directly by performing local least-squares curve fits in each cloud of points and then analytically differentiating the resulting curve fits to approximate the derivatives. A discussion of these topics is presented in viewgraph form.

John T Batina↗
NASA NTRS1988

Parameter estimation in nonlinear distributed systems - Approximation theory and convergence results

An abstract approximation framework and convergence theory is described for Galerkin approximations applied to inverse problems involving nonlinear distributed parameter systems. Parameter estimation problems are considered and formulated as the minimization of a least-squares-like performance index over a compact admissible parameter set subject to state constraints given by an inhomogeneous nonlinear distributed system. The theory applies to systems whose dynamics can be described by either time-independent or nonstationary strongly maximal monotonic operators defined on a reflexive Banach space which is densely and continuously embedded in a Hilbert space. It is demonstrated that if readily verifiable conditions on the system's dependence on the unknown parameters are satisfied, and the usual Galerkin approximation assumption holds, then solutions to the approximating problems exist and approximate a solution to the original infinite-dimensional identification problem.

Banks, H. T.↗
NASA NTRS1984

Fitting aerodynamic forces in the Laplace domain: An application of a nonlinear nongradient technique to multilevel constrained optimization

A technique which employs both linear and nonlinear methods in a multilevel optimization structure to best approximate generalized unsteady aerodynamic forces for arbitrary motion is described. Optimum selection of free parameters is made in a rational function approximation of the aerodynamic forces in the Laplace domain such that a best fit is obtained, in a least squares sense, to tabular data for purely oscillatory motion. The multilevel structure and the corresponding formulation of the objective models are presented which separate the reduction of the fit error into linear and nonlinear problems, thus enabling the use of linear methods where practical. Certain equality and inequality constraints that may be imposed are identified; a brief description of the nongradient, nonlinear optimizer which is used is given; and results which illustrate application of the method are presented.

Tiffany, S. H.↗
NASA NTRS2001

A Global Optimization Methodology for Rocket Propulsion Applications

While the response surface method is an effective method in engineering optimization, its accuracy is often affected by the use of limited amount of data points for model construction. In this chapter, the issues related to the accuracy of the RS approximations and possible ways of improving the RS model using appropriate treatments, including the iteratively re-weighted least square (IRLS) technique and the radial-basis neural networks, are investigated. A main interest is to identify ways to offer added capabilities for the RS method to be able to at least selectively improve the accuracy in regions of importance. An example is to target the high efficiency region of a fluid machinery design space so that the predictive power of the RS can be maximized when it matters most. Analytical models based on polynomials, with controlled level of noise, are used to assess the performance of these techniques.

Source record↗
DOE OSTI2025

Derivative-free stochastic optimization via adaptive sampling strategies

In this paper, we present a novel derivative-free framework for solving unconstrained stochastic optimization problems. Many problems in fields ranging from simulation optimization to reinforcement learning to quantum computing involve settings where only stochastic function values are obtained via a zeroth-order oracle, which has no available gradient information and necessitates the usage of derivative-free optimization methodologies. Our approach includes estimating gradients using stochastic function evaluations and integrating adaptive sampling techniques to control the accuracy in these stochastic approximations. Our framework encapsulates several gradient estimation techniques, including standard finite-difference, Gaussian smoothing, sphere smoothing, randomized coordinate finite-difference, and randomized subspace finite-difference methods. We provide theoretical convergence guarantees for our framework and analyze the worst-case iteration and sample complexities associated with each gradient estimation method. Finally, we demonstrate the empirical performance of the methods on logistic regression and nonlinear least squares problems.

Adaptive sampling↗