Search NASA⌕ Search

SEARCH · Search NASA

Results for “Identification problem”

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 559 records · Page 31

Extension and validation of a method for locating damaged members in large space trusses

The damage location approach employs the control system capabilities for the structure to test the structure and measure the dynamic response. The measurements are then used in a system identification algorithm to produce a model of the damaged structure. The model is compared to one for the undamaged structure to find regions of reduced stiffness which indicate the location of damage. Kabe's 3,4 stiffness matrix adjustment method was the central identification algorithm. The strength of his method is that, with minimal data, it preserves the representation of the physical connectivity of the structure in the resulting model of the damaged truss. However, extensive storage and computational effort were required as a result. Extension of the damage location method to overcome these problems is the first part of the current work. The central system identification algorithm is replaced with the MSMT method of stiffness matrix adjustment which was previously derived by generalizing an optimal-update secant method form quasi-Newton approaches for nonlinear optimization. Validation of the extended damage location method is the second goal.

Smith, Suzanne Weaver↗

Identification of subsurface structures using electromagnetic data and shape priors

We consider the inverse problem of identifying large-scale subsurface structures using the controlled source electromagnetic method. To identify structures in the subsurface where the contrast in electric conductivity can be small, regularization is needed to bias the solution towards preserving structural information. We propose to combine two approaches for regularization of the inverse problem. In the first approach we utilize a model-based, reduced, composite representation of the electric conductivity that is highly flexible, even for a moderate number of degrees of freedom. With a low number of parameters, the inverse problem is efficiently solved using a standard, second-order gradient-based optimization algorithm. Further regularization is obtained using structural prior information, available, e.g., from interpreted seismic data. The reduced conductivity representation is suitable for incorporation of structural prior information. Such prior information cannot, however, be accurately modeled with a gaussian distribution. To alleviate this, we incorporate the structural information using shape priors. The shape prior technique requires the choice of kernel function, which is application dependent. We argue for using the conditionally positive definite kernel which is shown to have computational advantages over the commonly applied gaussian kernel for our problem. Numerical experiments on various test cases show that the methodology is able to identify fairly complex subsurface electric conductivity distributions while preserving structural prior information during the inversion.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Parametric robust control and system identification: Unified approach

Despite significant advancement in the area of robust parametric control, the problem of synthesizing such a controller is still a wide open problem. Thus, we attempt to give a solution to this important problem. Our approach captures the parametric uncertainty as an H(sub infinity) unstructured uncertainty so that H(sub infinity) synthesis techniques are applicable. Although the techniques cannot cope with the exact parametric uncertainty, they give a reasonable guideline to model the unstructured uncertainty that contains the parametric uncertainty. An additional loop shaping technique is also introduced to relax its conservatism.

Keel, Leehyun↗

Scientific Discovery with Physics-Informed System Identification (Abbreviated Report)

My fellowship research focused on making physics-based simulations faster and more useful through machine learning. Many problems in science and engineering are governed by partial differential equations, but high-fidelity simulations are often too expensive to run repeatedly. I worked on improving Latent Space Dynamics Identification (LaSDI), a reduced-order modeling framework that compresses large simulation data sets into a smaller representation and then learns how that representation evolves over time. The motivation was to develop reduced models that remain accurate for more challenging systems, especially when predictions must remain reliable over long time intervals or when the underlying dynamics are more complicated than standard methods can easily handle. I also contributed to related work on Quandary, a high-performance software effort for simulation and control of open quantum systems, before focusing primarily on Latent Space Dynamics Identification methods. The main outcomes of the fellowship were two new algorithms (both of which were published), Rollout-LaSDI and Higher-Order LaSDI, together with supporting work on multi-stage Latent Space Dynamics Identification. Rollout-LaSDI improved long-term prediction by training the model to stay accurate over extended time horizons, and Higher-Order LaSDI broadened the method so it could model systems with higher-order time dynamics. My contributions to multistage Latent Space Dynamics Identification also helped show that its later training stages could be simplified without losing effectiveness, and that this behavior held across different model architectures and training strategies. Taken together, these advances improved the accuracy, flexibility, and practical value of reduced-order modeling tools for computational science.

97 MATHEMATICS AND COMPUTING↗

Latent Mechanisms of Polarization Switching from In Situ Electron Microscopy Observations

In situ scanning transmission electron microscopy enables observation of the domain dynamics in ferroelectric materials as a function of externally applied bias and temperature. The resultant data sets contain a wealth of information on polarization switching and phase transition mechanisms. However, identification of these mechanisms from observational data sets has remained a problem due to a large variety of possible configurations, many of which are degenerate. Here, an approach based on a combination of deep learning-based semantic segmentation, rotationally invariant variational autoencoder (VAE), and non-negative matrix factorization to enable learning of a latent space representation of the data with multiple real-space rotationally equivalent variants mapped to the same latent space descriptors is introduced. By varying the size of training sub-images in the VAE, the degree of complexity in the structural descriptors is tuned from simple domain wall detection to the identification of switching pathways. Importantly, this yields a powerful tool for the exploration of the dynamic data in mesoscopic electron, scanning probe, optical, and chemical imaging. Moreover, this work adds to the growing body of knowledge of incorporating physical constraints into the machine and deep-learning methods to improve learned descriptors of physical phenomena.

36 MATERIALS SCIENCE↗

An algorithm to correct saturated mass spectrometry ion abundances for enhanced quantitation and mass accuracy in omic studies

The mass accuracy and peak intensity of ions detected by mass spectrometry (MS) measurements are essential to facilitate compound identification and quantitation. However, high concentration species can easily cause problems if their ion intensities reach beyond the limits of the detection system, leading to distorted and non-ideal detector response (e.g. saturation), and largely precluding the calculation of accurate m/z and intensity values. Here we present an open source computational method to correct peaks above a defined intensity (saturated) threshold determined by the MS instrumentation such as the analog-to-digital converters or time-to-digital converters used in conjunction with time-of-flight MS. In this method, the isotopic envelope for each observed ion above the saturation threshold is compared to its expected theoretical isotopic distribution. The most intense isotopic peak for which saturation does not occur is then utilized to re-calculate the precursor m/z and correct the intensity, resulting in both higher mass accuracy and greater dynamic range. The benefits of this approach were evaluated with proteomic and lipidomic datasets of varying complexities. After correcting the high concentration species, reduced mass errors and enhanced dynamic range were observed for both simple and complex omic samples. Specifically, the mass error dropped by more than 50% in most cases with highly saturated species and dynamic range increased by 1-2 orders of magnitude for peptides in a blood serum sample.

Environmental Molecular Sciences Laboratory↗

Integrable perturbations of conformal field theories and Yetter-Drinfeld modules

In this paper we relate a problem in representation theory — the study of Yetter-Drinfeld modules over certain braided Hopf algebras — to a problem in two-dimensional quantum field theory, namely, the identification of integrable perturbations of a conformal field theory. A prescription that parallels Lusztig's construction allows one to read off the quantum group governing the integrable symmetry. As an example, we illustrate how the quantum group for the loop algebra of sl(2) appears in the integrable structure of the perturbed uncompactified and compactified free boson.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Non-autoregressive time-series methods for stable parametric reduced-order models

Advection-dominated dynamical systems, characterized by partial differential equations, are found in applications ranging from weather forecasting to engineering design where accuracy and robustness are crucial. There has been significant interest in the use of techniques borrowed from machine learning to reduce the computational expense and/or improve the accuracy of predictions for these systems. These rely on the identification of a basis that reduces the dimensionality of the problem and the subsequent use of time series and sequential learning methods to forecast the evolution of the reduced state. Often, however, machine-learned predictions after reduced-basis projection are plagued by issues of stability stemming from incomplete capture of multiscale processes as well as due to error growth for long forecast durations. To address these issues, we have developed a non-autoregressive time series approach for predicting linear reduced-basis time histories of forward models. In particular, we demonstrate that non-autoregressive counterparts of sequential learning methods such as long short-term memory (LSTM) considerably improve the stability of machine-learned reduced-order models. Further, we evaluate our approach on the inviscid shallow water equations and show that a non-autoregressive variant of the standard LSTM approach that is bidirectional in the principal component directions obtains the best accuracy for recreating the nonlinear dynamics of partial observations. Moreover-and critical for many applications of these surrogates-inference times are reduced by three orders of magnitude using our approach, compared with both the equation-based Galerkin projection method and the standard LSTM approach.

97 MATHEMATICS AND COMPUTING↗

STUDYING FREQUENCY RELATIONSHIPS OF KILOHERTZ QUASI-PERIODIC OSCILLATIONS FOR 4U 1636-53 AND Sco X-1: OBSERVATIONS CONFRONT THEORIES

By fitting the frequencies of simultaneous lower and upper kilohertz quasi-periodic oscillations (kHz QPOs) in two prototype neutron star (NS) QPO sources (4U 1636-53 and Sco X-1), we test the predictive power of all currently proposed QPO models. Models predict a linear, power law, or other relationship between the two frequencies. We found that for plausible NS parameters (mass and angular momentum), no model can satisfactorily reproduce the data, leading to very large chi-square values in our fittings. For both 4U 1636-53 and Sco X-1, this is largely due to the fact that the data significantly differ from a linear relationship. Some models perform relatively better but still have their own problems. Such a detailed comparison of data from models enables identification of routes for improving those models further.

79 ASTRONOMY AND ASTROPHYSICS↗

fPINNs: Fractional Physics-Informed Neural Networks

Physics-informed neural networks (PINNs), introduced, are effective in solving integer-order partial differential equations (PDEs) based on scattered and noisy data. PINNs employ standard feedforward neural networks (NNs) with the PDEs explicitly encoded into the NN using automatic differentiation, while the sum of the mean-squared PDE residuals and the mean-squared error in initial-boundary conditions is minimized with respect to the NN parameters. Here we extend PINNs to fractional PINNs (fPINNs) to solve space-time fractional advection-diffusion equations (fractional ADEs), and we study systematically their convergence, hence explaining both fPINNs and PINNs for the first time. Specifically, we demonstrate their accuracy and effectiveness in solving multidimensional forward and inverse problems with forcing terms whose values are only known at randomly scattered spatio-temporal coordinates (black-box (BB) forcing terms). A novel element of the fPINNs is the hybrid approach that we introduce for constructing the residual in the loss function using both automatic differentiation for the integer-order operators and numerical discretization for the fractional operators. This approach bypasses the difficulties stemming from the fact that automatic differentiation is not applicable to fractional operators because the standard chain rule in integer calculus is not valid in fractional calculus. To discretize the fractional operators, we employ the Grünwald--Letnikov (GL) formula in one-dimensional fractional ADEs and the vector GL formula in conjunction with the directional fractional Laplacian in two- and three-dimensional fractional ADEs. We first consider the one-dimensional fractional Poisson equation and compare the convergence of the fPINNs against the finite difference method (FDM). We present the solution convergence using both the mean L 2 error as well as the standard deviation due to sensitivity to NN parameter initializations. Using different GL formulas we observe first-, second-, and third-order convergence rates for small size training sets but the error saturates for larger training sets. We explain these results by analyzing the four sources of numerical errors due to discretization, sampling, NN approximation, and optimization. The total error decays monotonically (below 10 –5 for a third-order GL formula) but it saturates beyond that point due to the optimization error. We also analyze the relative balance between discretization and sampling errors and observe that the sampling size and the number of discretization points (auxiliary points) should be comparable to achieve the highest accuracy. As we increase the depth of the NN up to certain value, the mean error decreases and the standard deviation increases, whereas the width has essentially no effect unless its value is either too small or too large. We next consider time-dependent fractional ADEs and compare white-box (WB) and BB forcing. We observe that for the WB forcing, our results are similar to the aforementioned cases; however, for the BB forcing fPINNs outperform FDM. Subsequently, we consider multidimensional time-, space-, and space-time-fractional ADEs using the directional fractional Laplacian and we observe relative errors of 10 –3 ~ 10 –4 . Lastly, we solve several inverse problems in one, two, and three dimensions to identify the fractional orders, diffusion coefficients, and transport velocities and obtain accurate results given proper initializations even in the presence of significant noise.

97 MATHEMATICS AND COMPUTING↗

Experimental Mathematics and Mathematical Physics

One of the most effective techniques of experimental mathematics is to compute mathematical entities such as integrals, series or limits to high precision, then attempt to recognize the resulting numerical values. Recently these techniques have been applied with great success to problems in mathematical physics. Notable among these applications are the identification of some key multi-dimensional integrals that arise in Ising theory, quantum field theory and in magnetic spin theory.

97 MATHEMATICS AND COMPUTING↗

Weak Form Scientific Machine Learning: Test Function Construction for System Identification

Weak form Scientific Machine Learning (WSciML) is a recently developed framework for data-driven modeling and scientific discovery. It leverages the weak form of equation error residuals to provide enhanced noise robustness in system identification via convolving model equations with test functions, reformulating the problem to avoid direct differentiation of data. The performance, however, relies on wisely choosing a set of compactly supported test functions. In this work, we mathematically motivate a novel data-driven method for constructing Single-scale-Local reference functions for creating the set of test functions. Our approach numerically approximates the integration error introduced by the quadrature and identifies the support size for which the error is minimal, without requiring access to the model parameter values. Through numerical experiments across various models, noise levels, and temporal resolutions, we demonstrate that the selected supports consistently align with regions of minimal parameter estimation error. We also compare the proposed method against the strategy for constructing Multi-scale-Global (and orthogonal) test functions introduced in our prior work, demonstrating the improved computational efficiency.

FOS: Computer and information sciences↗

Recent advances in thermal-structural analysis and design

A continuing effort to develop a comprehensive capability for thermal-structural analysis and automated design (sizing) is reviewed. A principal role in the activity is played by the finite element program SPAR which contains both an efficient structural and thermal analysis capability. The benefit of having thermal and structural analyses in the same finite element program is illustrated by the application of SPAR to design calculations for the National Transonic Facility - a cryogenic wind tunnel under construction at Langley. Some experience with large-scale thermal structural analysis problems, particularly the space shuttle orbiter, has led to the identification of some analysis needs. Those needs include automated model generation and data output for lumped parameter thermal analysis, faster solution methods for nonlinear transient heat transfer, automated interpolation of temperature data from a thermal finite element model to a dissimilar structural finite model, and automated techniques to identify the times at which the critical combinations of transient heating and loads occur on a structure.

Adelman, H. M.↗

Lithium-thionyl chloride battery safety

Primary lithium cells which use LiAlCl4/SOCl2 electrolyte exhibit high energy density and long life. Currently these cells pose a safety problem since they have been found to vent or explode. This paper summarizes experiments to resolve the safety problem of Li-SOCl2 cells by thermal modeling and identification of possibly hazardous intermediates formed during discharge of these cells. A thermal model and mechanism for the reduction of SOCl2 are presented, as well as a discussion of their application to Li-SOCl2 cell safety.

Carter, B.↗

Structure Errors in System Identification

An approach to system identification is presented which explicitly takes structure errors into account and hence provides a systematic way for answering questions concerning the magnitude of estimated parameter errors resulting from structural errors. It is indicated that, from this point of view, it is possible to define near equivalence between process and model and to obtain meaningful theoretical results on solution error system identification. It remains to apply these results to large realistic problems such as those involving models of complex man machine systems.

Bekey, G. A.↗

WetNet: Using SSM/I data interactively for global distribution of tropical rainfall and precipitable water

The research objectives were the following: (1) to use SSM/I to categorize, measure, and parameterize effects of rainfall systems around the globe, especially mesoscale convective systems; (2) to use SSM/I to monitor key components of the global hydrologic cycle, including tropical rainfall and precipitable water, and links to increasing sea surface temperatures; and (3) to assist in the development of efficient methods of exchange of massive satellite data bases and of analysis techniques, especially their use at a university. Numerous tasks have been initiated. First and foremost has been the integration and startup of the WetNet computer system into the TAMU computer network. Scientific activity was infeasible before completion of this activity. Final hardware delivery was not completed until October 1991, after which followed a period of identification and solution of several hardware and software and software problems. Accomplishments representing approximately four months work with the WetNEt system are presented.

Zipser, Edward J.↗

Human-Computer Interaction with Medical Decisions Support Systems

Decision Support Systems (DSSs) have been available to medical diagnosticians for some time, yet their acceptance and use have not increased with advances in technology and availability of DSS tools. Medical DSSs will be necessary on future long duration space missions, because access to medical resources and personnel will be limited. Human-Computer Interaction (HCI) experts at NASA's Human Factors and Ergonomics Laboratory (HFEL) have been working toward understanding how humans use DSSs, with the goal of being able to identify and solve the problems associated with these systems. Work to date consists of identification of HCI research areas, development of a decision making model, and completion of two experiments dealing with 'anchoring'. Anchoring is a phenomenon in which the decision maker latches on to a starting point and does not make sufficient adjustments when new data are presented. HFEL personnel have replicated a well-known anchoring experiment and have investigated the effects of user level of knowledge. Future work includes further experimentation on level of knowledge, confidence in the source of information and sequential decision making.

Adolf, Jurine A.↗

GRIDS-Net: Inverse shape design and identification of scatterers via geometric regularization and physics-embedded deep learning

This study presents a deep learning based methodology for both remote sensing and design of acoustic scatterers. The ability to determine the shape of a scatterer, either in the context of material design or sensing, plays a critical role in many practical engineering problems. This class of inverse problems is extremely challenging due to their high-dimensional, nonlinear, and ill-posed nature. To overcome these technical hurdles, we introduce a geometric regularization approach for deep neural networks (DNN) based on non-uniform rational B-splines (NURBS) and capable of predicting complex 2D scatterer geometries in a parsimonious dimensional representation. Then, this geometric regularization is combined with physics-embedded learning and integrated within a robust convolutional autoencoder (CAE) architecture to accurately predict the shape of 2D scatterers in the context of identification and inverse design problems. Further, an extensive numerical study is presented in order to showcase the remarkable ability of this approach to handle complex scatterer geometries while generating physically-consistent acoustic fields. The study also assesses and contrasts the role played by the (weakly) embedded physics in the convergence of the DNN predictions to a physically consistent inverse design.

42 ENGINEERING↗