Search NASA⌕ Search

SEARCH · Search NASA

Results for “extended least squares”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
NASA NTRS1995

A concept for adaptive performance optimization on commercial transport aircraft

An adaptive control method is presented for the minimization of drag during flight for transport aircraft. The minimization of drag is achieved by taking advantage of the redundant control capability available in the pitch axis, with the horizontal tail used as the primary surface and symmetric deflection of the ailerons and cruise flaps used as additional controls. The additional control surfaces are excited with sinusoidal signals, while the altitude and velocity loops are closed with guidance and control laws. A model of the throttle response as a function of the additional control surfaces is formulated and the parameters in the model are estimated from the sensor measurements using a least squares estimation method. The estimated model is used to determine the minimum drag positions of the control surfaces. The method is presented for the optimization of one and two additional control surfaces. The adaptive control method is extended to optimize rate of climb with the throttle fixed. Simulations that include realistic disturbances are presented, as well as the results of a Monte Carlo simulation analysis that shows the effects of changing the disturbance environment and the excitation signal parameters.

Jackson, Michael R.↗
NASA NTRS2002

Kernel Partial Least Squares for Nonlinear Regression and Discrimination

This paper summarizes recent results on applying the method of partial least squares (PLS) in a reproducing kernel Hilbert space (RKHS). A previously proposed kernel PLS regression model was proven to be competitive with other regularized regression methods in RKHS. The family of nonlinear kernel-based PLS models is extended by considering the kernel PLS method for discrimination. Theoretical and experimental results on a two-class discrimination problem indicate usefulness of the method.

Rosipal, Roman↗
DOE OSTI2026

ReMU: regional minimal updating for model-based derivative-free optimization

Derivative-free optimization (DFO) problems are optimization problems where derivative information is unavailable or extremely difficult to obtain. Model-based DFO solvers have been applied extensively in scientific computing. Powell's NEWUOA (2004) [Powell, The NEWUOA software for unconstrained optimization without derivatives, in Large-Scale Nonlinear Optimization, Nonconvex Optimization and its Applications Vol. 83, G. Di Pillo and M. Roma, eds., Springer, 2006, pp. 255–297] and Wild's POUNDerS (2014) [Wild, Solving derivative-free nonlinear least squares problems with POUNDERS, in Advances and Trends in Optimization with Engineering Applications, T. Terlaky, M.F. Anjos, and S. Ahmed, eds., SIAM, 2017, pp. 529–540] explore the numerical power of the minimal norm Hessian (MNH) model for DFO and contributed to the open discussion on building better models with fewer data to achieve faster numerical convergence. Another decade later, we propose the regional minimal updating (ReMU) models, and extend the previous models into a broader class, including the H 2 norm models [Xie and Yuan, Least H 2 norm updating of quadratic interpolation models for derivative-free trust-region algorithms, IMA J. Numer. Anal. 46 (2025), pp. 21–50]. This paper shows motivation behind ReMU models, computational details, theoretical and numerical results on particular extreme points and the barycentre of ReMU's weight coefficient region, and the associated KKT matrix error and distance. Novel metrics, such as the truncated Newton step error, are proposed to numerically understand the new models' properties. A new algorithmic strategy, based on iteratively adjusting the ReMU model type, is also proposed, and shows numerical advantages by combining and switching between the barycentric model and the classic least Frobenius norm model in an online fashion.

derivative-free trust-region methods↗
DOE OSTI2022

Artificial intelligence based analysis of nanoindentation load–displacement data using a genetic algorithm

In this work, we developed an automated tool, Nanoindentation Neo package for the analysis of nanoindentation load–displacement curves using a Genetic Algorithm (GA) applied to the Oliver-Pharr method (Oliver et al.,1992). For some materials, such as polycrystalline isotropic graphites, Least Squares Fitting (LSF) of the unload curve can produce unrealistic fit parameters. These graphites exhibit sharply peaked unloading curves not easily fit using the LSF, which tends to overestimate the indenter tip geometry parameter. To tackle this problem, we extended our general materials characterization tool Neo for EXAFS analysis (Terry et al., 2021) to fit nanoindentation data. Nanoindentation Neo automatically processes and analyzes nanoindentation data with minimal user input while producing meaningful fit parameters. GA, a robust metaheuristic method, begins with a population of temporary solutions using model parameters called chromosomes; from these we evaluate a fitness value for each solution, and select the best solutions to mix with random solutions producing the next generation. A mutation operator then modifies existing solutions by random perturbations, and the optimal solution is selected. We tested the GA method using Silica and Al reference standards. We fit samples of graphite and a high entropy alloy (HEA) consisting of BCC and FCC phases.

42 ENGINEERING↗
DOE OSTI2023

Open-Shell Tensor Hypercontraction

The extension of least-squares tensor hypercontracted second- and third-order Møller–Plesset perturbation theory (LS-THC-MP2 and LS-THC-MP3) to open-shell systems is an important development due to the scaling reduction afforded by THC and the ubiquity of molecular ions, radicals, and other open-shell reactive species. The complexity of wavefunction-based quantum-chemical methods such as Møller–Plesset and coupled cluster theory is reflected in the steep scaling of the computational costs with the molecular size. The least-squares tensor hypercontraction (LS-THC) method is an efficient, single-step factorization for the two-electron integral tensor but can also be used to factorize the double excitation amplitudes, leading to significant scaling reduction. Herein we extend this promising method to open-shell variants of LS-THC-MP2 and -MP3 by using diagrammatic techniques and explicit spin summation. The accuracy of the resulting methods for open-shell species is benchmarked on standard test systems such as regular alkanes as well as realistic systems involving bond breaking, radical stabilization, and other effects. We find that open-shell LS-THC-MPn methods exhibit errors highly comparable to those produced by closed-shell LS-THC-MPn and are highly insensitive to particular chemical interactions, geometries, or even moderate spin contamination.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Accelerating gradient descent and Adam via fractional gradients

Here we propose a class of novel fractional-order optimization algorithms. We define a fractional-order gradient via the Caputo fractional derivatives that generalizes integer-order gradient. We refer it to as the Caputo fractional-based gradient, and develop an efficient implementation to compute it. A general class of fractional-order optimization methods is then obtained by replacing integer-order gradients with the Caputo fractional-based gradients. To give concrete algorithms, we consider gradient descent (GD) and Adam, and extend them to the Caputo fractional GD (CfGD) and the Caputo fractional Adam (CfAdam). We demonstrate the superiority of CfGD and CfAdam on several large scale optimization problems that arise from scientific machine learning applications, such as ill-conditioned least squares problem on real-world data and the training of neural networks involving non-convex objective functions. Numerical examples show that both CfGD and CfAdam result in acceleration over GD and Adam, respectively. We also derive error bounds of CfGD for quadratic functions, which further indicate that CfGD could mitigate the dependence on the condition number in the rate of convergence and results in significant acceleration over GD.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2017

Reconfigurable Cellular Composite Structures for Lighter than Air Vehicles with Scalable Size and Endurance

Engineered non-stochastic cellular materials show promising characteristics on the laboratory scale,with nearly ideal specific stiffness and strength scaling at ultralight mass density. These propertiessuggest performance benefits in any application with combined stiffness and mass constraints, suchas air vehicles. We investigate here the application of re-configurable cellular composite materialsand structures to lighter than air vehicles. We describe the properties and applicability of these materials,provide an example analysis of governing loading conditions associated with airships, showan example optimization method for navigating the design space, and describe how recent advancesin cellular material manufacturing and reconfiguration enable system performance benefits includingnew concepts of operation. Lastly, we propose lighter than air vehicles that are assembled andmaintained in-flight, eliminating structural compromises associated with transitional flight modesand ground handling.Engineered non-stochastic cellular material properties suggest performance benefits in lighter than air vehicles due tostiffness and mass constraints that are intrinsic to the airship design problem. Recent advances in cellular materialmanufacturing and reconfiguration enable system performance benefits including new concepts of operation, such aslighter than air vehicles that are assembled and maintained in-flight, eliminating structural compromises associatedwith transitional flight modes and ground handling. Existing engineered cellular materials display properties allowinglarge large scale airships design as monocoque cellular solids. Inevitable improvements in cellular material propertiesand manufacturing will improve feasibility even further. Given the suggestion that the two most significant technologygaps exist across all current airship projects are manufacturing and assembly processes and ground handling [7],a strategy that encompasses construction and maintenance in flight could provide critical rephrasing of the systemdesign problem through these new concepts of operation. Refactoring of traditional manufacturing, operation, andservice process constraints could extend to other domains in aerospace systems and manufacturing in general.In future work, the complexity of the design task would benefit from a form of optimization in order to find themost suitable geometry for a chosen application. For example, the Sequential Least SQuares Programming (SLSQP)function from within the SciPy Minimize library is a multiobjective constrained optimization method that has beenapplied to fixed wing aircraft design. [17] In this situation it would allow for several objective functions such as drag,bending stiffness, buoyancy and cost of transport to be incorporated into a composite objective function.

Copplestone, Grace↗
NASA NTRS1975

Sequential Least-Squares Using Orthogonal Transformations

Square root information estimation, starting from its beginnings in least-squares parameter estimation, is considered. Special attention is devoted to discussions of sensitivity and perturbation matrices, computed solutions and their formal statistics, consider-parameters and consider-covariances, and the effects of a priori statistics. The constant-parameter model is extended to include time-varying parameters and process noise, and the error analysis capabilities are generalized. Efficient and elegant smoothing results are obtained as easy consequences of the filter formulation. The value of the techniques is demonstrated by the navigation results that were obtained for the Mariner Venus-Mercury (Mariner 10) multiple-planetary space probe and for the Viking Mars space mission.

Bierman, G. J.↗
NASA NTRS1991

Image construction from the IRAS survey

IRAS survey data can be used successfully to produce images of extended objects. The major difficulties, i.e., nonuniform sampling, different response functions for each detector, and varying signal-to-noise levels for each detector for each scan, have been resolved. The results of three different image construction techniques are compared: coaddition, constrained least-squares, and maximum entropy. The maximum entropy result is superior. An image of the galaxy M51 with an average spatial resolution of 45 arcsecods, using 60 microns survey data is presented. This exceeds the telescope diffraction limit of about 1 minute of arc at this wavelength.

Bontekoe, T. R.↗
NASA NTRS2001

A {1,2}-Order Plate Theory Accounting for Three-Dimensional Thermoelastic Deformations in Thick Composite and Sandwich Laminates

A {1,2}-order theory for laminated composite and sandwich plates is extended to include thermoelastic effects. The theory incorporates all three-dimensional strains and stresses. Mixed-field assumptions are introduced which include linear in-plane displacements, parabolic transverse displacement and shear strains, and a cubic distribution of the transverse normal stress. Least squares strain compatibility conditions and exact traction boundary conditions are enforced to yield higher polynomial degree distributions for the transverse shear strains and transverse normal stress through the plate thickness. The principle of virtual work is used to derive a 10th-order system of equilibrium equations and associated Poisson boundary conditions. The predictive capability of the theory is demonstrated using a closed-form analytic solution for a simply-supported rectangular plate subjected to a linearly varying temperature field across the thickness. Several thin and moderately thick laminated composite and sandwich plates are analyzed. Numerical comparisons are made with corresponding solutions of the first-order shear deformation theory and three-dimensional elasticity theory. These results, which closely approximate the three-dimensional elasticity solutions, demonstrate that through - the - thickness deformations even in relatively thin and, especially in thick. composite and sandwich laminates can be significant under severe thermal gradients. The {1,2}-order kinematic assumptions insure an overall accurate theory that is in general superior and, in some cases, equivalent to the first-order theory.

Tessler, A.↗
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↗
NASA NTRS1990

Image reconstruction of IRAS survey scans

The IRAS survey data can be used successfully to produce images of extended objects. The major difficulties, viz. non-uniform sampling, different response functions for each detector, and varying signal-to-noise levels for each detector for each scan, were resolved. The results of three different image construction techniques are compared: co-addition, constrained least squares, and maximum entropy. The maximum entropy result is superior. An image of the galaxy M51 with an average spatial resolution of 45 arc seconds is presented, using 60 micron survey data. This exceeds the telescope diffraction limit of 1 minute of arc, at this wavelength. Data fusion is a proposed method for combining data from different instruments, with different spacial resolutions, at different wavelengths. Data estimates of the physical parameters, temperature, density and composition, can be made from the data without prior image (re-)construction. An increase in the accuracy of these parameters is expected as the result of this more systematic approach.

Bontekoe, Tj. Romke↗
NASA NTRS1990

Image construction from the IRAS survey and data fusion

The IRAS survey data can be used successfully to produce images of extended objects. The major difficulty, viz. non-uniform sampling, different response functions for each detector, and varying signal-to-noise levels for each detector for each scan, were resolved. The results of three different image construction techniques are compared: co-addition, constrained least squares, and maximum entropy. The maximum entropy result is superior. An image of the galaxy M51 with an average spatial resolution of 45 arc seconds, is presented using 60 micron survey data. This exceeds the telescope diffraction limit of 1 minute of arc, at this wavelength. Data fusion is a proposed method for combining data from different instruments, with different spatial resolutions, at different wavelengths. Direct estimates of the physical parameters, temperature, density and composition, can be made from the data without prior images (re-)construction. An increase in the accuracy of these parameters is expected as the result of this more systematic approach.

Bontekoe, Tj. R.↗
NASA NTRS2011

Aperiodicity Correction for Rotor Tip Vortex Measurements

The initial roll-up of a tip vortex trailing from a model-scale, hovering rotor was measured using particle image velocimetry. The unique feature of the measurements was that a microscope was attached to the camera to allow much higher spatial resolution than hitherto possible. This also posed some unique challenges. In particular, the existing methodologies to correct for aperiodicity in the tip vortex locations could not be easily extended to the present measurements. The difficulty stemmed from the inability to accurately determine the vortex center, which is a prerequisite for the correction procedure. A new method is proposed for determining the vortex center, as well as the vortex core properties, using a least-squares fit approach. This approach has the obvious advantage that the properties are derived from not just a few points near the vortex core, but from a much larger area of flow measurements. Results clearly demonstrate the advantage in the form of reduced variation in the estimated core properties, and also the self-consistent results obtained using three different aperiodicity correction methods.

Ramasamy, Manikandan↗
DOE OSTI2020

Operando structural and chemical evolutions of TiS 2 in Na-ion batteries

Titanium disulfide (TiS 2 ) with high electric conductivity, fast rate capability, and good cycling performance is a promising candidate for electrode material in sodium (Na)-ion batteries. Despite the well-studied electrochemical behaviors of TiS 2 in Li-ion batteries, the detailed reaction mechanism of TiS 2 in Na-ion batteries is not yet fully understood due to a more complex multi-phase conversion process. In this work, reactions of TiS 2 in Na-ion batteries are investigated via a multi-modal synchrotron approach: operando X-ray Absorption Spectroscopy (XAS) – including X-ray Absorption Near-Edge Structure (XANES) and Extended X-ray Absorption Fine Structure (EXAFS) – and ex situ X-ray Powder Diffraction (XPD), coupled with computational modeling. Operando XANES spectra indicate that the redox reactions occur in both Ti and S during the electrochemically driven phase transformation. Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) analysis of XAS suggests that different numbers of components are involved in the lithiation and sodiation of TiS 2 , with the sodiation including at least one intermediate phase in addition to the starting material and the final sodiation product. Ex situ XPD and Rietveld refinement further determined and quantified the unknown phases, showing that three phases, TiS 2 , Na 0.55 TiS 2 , and NaTiS 2 , participate in the sodiation of TiS 2 . Operando EXAFS results show the changes in the Ti–Ti coordination number and interatomic distance. This explains the coulombic efficiency decay due to the incomplete recovery of the coordination number of Ti after cycling. Overall, this work reveals the reaction mechanism occurring in Na–TiS 2 batteries with a greater quantitative understanding of the structural evolution. By combining the multi-modal synchrotron approach and computational work, this study provides a framework for studying a broader range of electrochemically driven phase-transformation systems towards advanced energy storage and conversion applications.

25 ENERGY STORAGE↗
DOE OSTI2021

The Area Localized Coupled Model for Analytical Mean Flow Prediction in Arbitrary Wind Farm Geometries

This work introduces the area localized coupled (ALC) model, which extends the applicability of approaches that couple classical wake superposition models and atmospheric boundary layer models to wind farms with arbitrary layouts. Coupling wake and top–down boundary layer models is particularly challenging since the latter requires averaging over planform areas associated with turbine-specific regions of the flow that need to be specified. The ALC model uses Voronoi tessellation to define this local area around each turbine. A top–down description of a developing internal boundary layer is then applied over Voronoi cells upstream of each turbine to estimate the local mean velocity profile. Coupling between the velocity at hub-height based on this localized top–down model and a wake model is achieved by enforcing a minimum least-square-error in mean velocity in each cell. The wake model in the present implementation takes into account variations in wind farm inflow velocity and represents the wake profile behind each turbine as a super-Gaussian function that smoothly transitions between a top-hat shape in the region immediately following the turbine to a Gaussian profile downstream. Detailed comparisons to large-eddy simulation (LES) data from two different wind farms demonstrate the efficacy of the model in accurately predicting both wind farm power output and local turbine hub-height velocity for different wind farm geometries. These validations using data generated from two different LES codes demonstrate the model's versatility with respect to capturing results from different simulation setups and wind farm configurations.

49 EE - Wind and Water Power Program - Wind (EE-4W↗
DOE OSTI2025

Remote Americium Detection Using an Optical Sensor: A D-Optimal Strategy for Efficient PLS-Based Modeling

A fiber-optic visible–near-infrared absorption spectroscopy system in a glove box was demonstrated for remote quantification of Am(III) (0–500 µM) and HNO 3 (0.1–9 M) using partial least squares regression (PLSR) models. The sensor platform, featuring a simple plug-and-play spectrophotometer, can enable noninvasive, real-time monitoring of actinide process solutions. To establish a flexible PLSR model calibration strategy, a D-optimal design developed using Nd(III) in previous studies was successfully extended to an actinide system with Am(III) to effectively minimize sample set size while maintaining robust prediction performance. The results suggest strong spectral similarities between Nd(III) and Am(III) and validate Nd(III) as an effective optical surrogate for trivalent actinide species. This work also supports the generalizability of a D-optimal training set selection approach for two-factor systems. The PLS1 models for Am(III) and HNO 3 outperformed a PLS2 model and maintained reasonable performance in the presence of interfering U(VI). The resulting sensor system and multivariate approach provides a flexible and scalable solution for process monitoring, control, and safety in diverse nuclear applications.

actinide↗
NASA NTRS2014

Optimal Energy Measurement in Nonlinear Systems: An Application of Differential Geometry

Design of TES microcalorimeters requires a tradeoff between resolution and dynamic range. Often, experimenters will require linearity for the highest energy signals, which requires additional heat capacity be added to the detector. This results in a reduction of low energy resolution in the detector. We derive and demonstrate an algorithm that allows operation far into the nonlinear regime with little loss in spectral resolution. We use a least squares optimal filter that varies with photon energy to accommodate the nonlinearity of the detector and the non-stationarity of the noise. The fitting process we use can be seen as an application of differential geometry. This recognition provides a set of well-developed tools to extend our work to more complex situations. The proper calibration of a nonlinear microcalorimeter requires a source with densely spaced narrow lines. A pulsed laser multi-photon source is used here, and is seen to be a powerful tool for allowing us to develop practical systems with significant detector nonlinearity. The combination of our analysis techniques and the multi-photon laser source create a powerful tool for increasing the performance of future TES microcalorimeters.

Nonlinear response↗