Search NASA⌕ Search

SEARCH · Search NASA

Results for “Inverse 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 721 records · Page 40

An adaptive and stability-promoting layerwise training approach for sparse deep neural network architecture

This work presents a two-stage adaptive framework for progressively developing deep neural network (DNN) architectures that generalize well for a given training data set. In the first stage, a layerwise training approach is adopted where a new layer is added each time and trained independently by freezing parameters in the previous layers. We impose desirable structures on the DNN by employing manifold regularization, sparsity regularization, and physics-informed terms. We introduce a ε – δ – stability-promoting concept as a desirable property for a learning algorithm and show that employing manifold regularization yields a ε – δ stability-promoting algorithm. Further, we also derive the necessary conditions for the trainability of a newly added layer and investigate the training saturation problem. In the second stage of the algorithm (post-processing), a sequence of shallow networks is employed to extract information from the residual produced in the first stage, thereby improving the prediction accuracy. Numerical investigations on prototype regression and classification problems demonstrate that the proposed approach can outperform fully connected DNNs of the same size. Moreover, by equipping the physics-informed neural network (PINN) with the proposed adaptive architecture strategy to solve partial differential equations, we numerically show that adaptive PINNs not only are superior to standard PINNs but also produce interpretable hidden layers with provable stability. As a result, we also apply our architecture design strategy to solve inverse problems governed by elliptic partial differential equations.

42 ENGINEERING↗

STRATEGIES IN SEISMIC INFERENCE OF SUPERGRANULAR FLOWS ON THE SUN

Observations of the solar surface reveal the presence of flows with length scales of around 35 Mm, commonly referred to as supergranules. Inferring the subsurface flow profile of supergranules from measurements of the surface and photospheric wavefield is an important challenge faced by helioseismology. Traditionally, the inverse problem has been approached by studying the linear response of seismic waves in a horizontally translationally invariant background to the presence of the supergranule; following an iterative approach that does not depend on horizontal translational invariance might perform better, since the misfit can be analyzed post iterations. In this work, we construct synthetic observations using a reference supergranule and invert for the flow profile using surface measurements of travel times of waves belonging to modal ridges f (surface gravity) and p {sub 1} through p {sub 7} (acoustic). We study the extent to which individual modes and their combinations contribute to infer the flow. We show that this method of nonlinear iterative inversion tends to underestimate the flow velocities, as well as inferring a shallower flow profile, with significant deviations from the reference supergranule near the surface. We carry out a similar analysis for a sound-speed perturbation and find that analogous near-surface deviations persist, although the iterations converge faster and more accurately. We conclude that a better approach to inversion would be to expand the supergranule profile in an appropriate basis, thereby reducing the number of parameters being inverted for and appropriately regularizing them.

79 ASTRONOMY AND ASTROPHYSICS↗

A Comparison of Three Ways to Measure Time-Dependent Densities With Quantum Simulators

Quantum algorithms are touted as a way around some classically intractable problems such as the simulation of quantum mechanics. At the end of all quantum algorithms is a quantum measurement whereby classical data is extracted and utilized. In fact, many of the modern hybrid-classical approaches are essentially quantum measurements of states with short quantum circuit descriptions. Here, we compare and examine three methods of extracting the time-dependent one-particle probability density from a quantum simulation: direct Z -measurement, Bayesian phase estimation, and harmonic inversion. We have tested these methods in the context of the potential inversion problem of time-dependent density functional theory. Our test results suggest that direct measurement is the preferable method. We also highlight areas where the other two methods may be useful and report on tests using Rigetti's quantum virtual device. This study provides a starting point for imminent applications of quantum computing.

Yang, Jun↗

A fast, matrix-based method to perform omnidirectional pressure integration

Abstract Experimentally-measured pressure fields play an important role in understanding many fluid dynamics problems. Unfortunately, pressure fields are difficult to measure directly with non-invasive, spatially resolved diagnostics, and calculations of pressure from velocity have proven sensitive to error in the data. Omnidirectional line integration methods are usually more accurate and robust to these effects as compared to implicit Poisson equations, but have seen slower uptake due to the higher computational and memory costs, particularly in 3D domains. This paper demonstrates how omnidirectional line integration approaches can be converted to a matrix inversion problem. This novel formulation uses an iterative approach so that the boundary conditions are updated each step, preserving the convergence behavior of omnidirectional schemes while also keeping the computational efficiency of Poisson solvers. This method is implemented in Matlab and also as a GPU-accelerated code in CUDA-C++. The behavior of the new method is demonstrated on 2D and 3D synthetic and experimental data. Three-dimensional grid sizes of up to 125 million grid points are tractable with this method, opening exciting opportunities to perform volumetric pressure field estimation from 3D PIV measurements.

42 ENGINEERING↗

Reveal, A General Reverse Engineering Algorithm for Inference of Genetic Network Architectures

Given the immanent gene expression mapping covering whole genomes during development, health and disease, we seek computational methods to maximize functional inference from such large data sets. Is it possible, in principle, to completely infer a complex regulatory network architecture from input/output patterns of its variables? We investigated this possibility using binary models of genetic networks. Trajectories, or state transition tables of Boolean nets, resemble time series of gene expression. By systematically analyzing the mutual information between input states and output states, one is able to infer the sets of input elements controlling each element or gene in the network. This process is unequivocal and exact for complete state transition tables. We implemented this REVerse Engineering ALgorithm (REVEAL) in a C program, and found the problem to be tractable within the conditions tested so far. For n = 50 (elements) and k = 3 (inputs per element), the analysis of incomplete state transition tables (100 state transition pairs out of a possible 10(exp 15)) reliably produced the original rule and wiring sets. While this study is limited to synchronous Boolean networks, the algorithm is generalizable to include multi-state models, essentially allowing direct application to realistic biological data sets. The ability to adequately solve the inverse problem may enable in-depth analysis of complex dynamic systems in biology and other fields.

Liang, Shoudan↗

Parallel Solver Framework for Mixed-Integer PDE-Constrained Optimization

ROL-PEBBL is a C++, MPI-based parallel code for mixed-integer PDE-constrained optimization (MIPDECO). In these problems we wish to optimize (control, design, etc.) physical systems, which must obey the laws of physics, when some of the decision variables must take integer values. ROL-PEBBL combines a code to efficiently search over integer choices (PEBBL = Parallel Enumeration Branch-and-Bound Library) and a code for efficient nonlinear optimization, including PDE-constrained optimization (ROL = Rapid Optimization Library). In this report, we summarize the design of ROL-PEBBL and initial applications/results. For an artificial source-inversion problem, finding sources of pollution on a grid from sparse samples, ROL-PEBBLs solution for the nest grid gave the best optimization guarantee for any general solver that gives both a solution and a quality guarantee.

97 MATHEMATICS AND COMPUTING↗

Interpretation of solar extinction data for stratospheric aerosols

This paper discusses the inversion problem for aerosols using the solar extinction method. A series of numerical experiments is described in which solar extinction measurement systems are modeled. A numerical model of a solar extinction measurement system has been coupled with model atmospheres that exhibit fine scale structures to produce numerically generated data signals. These signals were then inverted to study the effect that measurement errors and desired vertical resolution produce in the inverted results. Knowledge o2 the trade off between vertical resolution and the accuracy of inversion aid in the interpretation of the inverted results.

Pepin, T. J.↗

Bayesian operator inference for data-driven reduced-order modeling

This work proposes a Bayesian inference method for the reduced-order modeling of time-dependent systems. Informed by the structure of the governing equations, the task of learning a reduced-order model from data is posed as a Bayesian inverse problem with Gaussian prior and likelihood. The resulting posterior distribution characterizes the operators defining the reduced-order model, hence the predictions subsequently issued by the reduced-order model are endowed with uncertainty. The statistical moments of these predictions are estimated via a Monte Carlo sampling of the posterior distribution. Since the reduced models are fast to solve, this sampling is computationally efficient. Furthermore, the proposed Bayesian framework provides a statistical interpretation of the regularization term that is present in the deterministic operator inference problem, and the empirical Bayes approach of maximum marginal likelihood suggests a selection algorithm for the regularization hyperparameters. The proposed method is demonstrated on two examples: the compressible Euler equations with noise-corrupted observations, and a single-injector combustion process.

97 MATHEMATICS AND COMPUTING↗

Numerical identification of boundary conditions on nonlinearly radiating inverse heat conduction problems

An explicit and unconditionally stable finite difference method for the solution of the transient inverse heat conduction problem in a semi-infinite or finite slab mediums subject to nonlinear radiation boundary conditions is presented. After measuring two interior temperature histories, the mollification method is used to determine the surface transient heat source if the energy radiation law is known. Alternatively, if the active surface is heated by a source at a rate proportional to a given function, the nonlinear surface radiation law is then recovered as a function of the interface temperature when the problem is feasible. Two typical examples corresponding to Newton cooling law and Stefan-Boltzmann radiation law respectively are illustrated. In all cases, the method predicts the surface conditions with an accuracy suitable for many practical purposes.

Murio, Diego A.↗

On Characterizing Particle Shape

It is well known that particle shape affects flow characteristics of granular materials, as well as a variety of other solids processing issues such as compaction, rheology, filtration and other two-phase flow problems. The impact of shape crosses many diverse and commercially important applications, including pharmaceuticals, civil engineering, metallurgy, health, and food processing. Two applications studied here include the dry solids flow of lunar simulants (e.g. JSC-1, NU-LHT-2M, OB-1), and the flow properties of wet concrete, including final compressive strength. A multi-dimensional generalized, engineering method to quantitatively characterize particle shapes has been developed, applicable to both single particle orientation and multi-particle assemblies. The two-dimension, three dimension inversion problem is also treated, and the application of these methods to DEM model particles will be discussed. In the case of lunar simulants, flow properties of six lunar simulants have been measured, and the impact of particle shape on flowability - as characterized by the shape method developed here -- is discussed, especially in the context of three simulants of similar size range. In the context of concrete processing, concrete construction is a major contributor to greenhouse gas production, of which the major contributor is cement binding loading. Any optimization in concrete rheology and packing that can reduce cement loading and improve strength loading can also reduce currently required construction safety factors. The characterization approach here is also demonstrated for the impact of rock aggregate shape on concrete slump rheology and dry compressive strength.

Ennis, Bryan J.↗

Inferring effective electrostatic interaction of charge-stabilized colloids from scattering using deep learning

In this article, an innovative strategy is presented that incorporates deep auto-encoder networks into a least-squares fitting framework to address the potential inversion problem in small-angle scattering. To evaluate the performance of the proposed approach, a detailed case study focusing on charged colloidal suspensions was carried out. The results clearly indicate that a deep learning solution offers a reliable and quantitative method for studying molecular interactions. The approach surpasses existing deterministic approaches with respect to both numerical accuracy and computational efficiency. Overall, this work demonstrates the potential of deep learning techniques in tackling complex problems in soft-matter structures and beyond.

36 MATERIALS SCIENCE↗

Phase Imaging: A Compressive Sensing Approach

Since Wolfgang Pauli posed the question in 1933, whether the probability densities |Ψ(r)|² (real-space image) and |Ψ(q)|² (reciprocal space image) uniquely determine the wave function Ψ(r) [1], the so called Pauli Problem sparked numerous methods in all fields of microscopy [2, 3]. Reconstructing the complete wave function Ψ(r) = a(r)e-iφ(r) with the amplitude a(r) and the phase φ(r) from the recorded intensity enables the possibility to directly study the electric and magnetic properties of the sample through the phase. In transmission electron microscopy (TEM), electron holography is by far the most established method for phase reconstruction [4]. Requiring a high stability of the microscope, next to the installation of a biprism in the TEM, holography cannot be applied to any microscope straightforwardly. Recently, a phase retrieval approach was proposed using conventional TEM electron diffractive imaging (EDI). Using the SAD aperture as reciprocal-space constraint, a localized sample structure can be reconstructed from its diffraction pattern and a real-space image using the hybrid input-output algorithm [5]. We present an alternative approach using compressive phase-retrieval [6]. Our approach does not require a real-space image. Instead, random complimentary pairs of checkerboard masks are cut into a 200 nm Pt foil covering a conventional TEM aperture (cf. Figure 1). Used as SAD aperture, subsequently diffraction patterns are recorded from the same sample area. Hereby every mask blocks different parts of gold particles on a carbon support (cf. Figure 2). The compressive sensing problem has the following formulation. First, we note that the complex-valued reciprocal-space wave-function is the Fourier transform of the (also complex-valued) real-space wave-function, Ψ(q) = F[Ψ(r)], and subsequently the diffraction pattern image is given by |Ψ(q)|2 = |F[Ψ(r)]|2. We want to find Ψ(r) given a few differently coded diffraction pattern measurements yn = |F[HnΨ(r)]|2, where the matrices Hn encode the mask structure of the aperture. This is a nonlinear inverse problem, but has been shown to be solvable even in the underdetermined case [6]. Since each diffraction pattern yn contains diffraction information from selected regions of the same sample, the differences in each pattern contain local phase information, which can be combined to form a full estimate of the real-space wave-function[7]. References: [1] W. Pauli in “Die allgemeinen Prinzipien der Wellenmechanik“, ed. H Geiger and W Scheel, (Julius Springer, Berlin). [2] A. Tonomura, Rev. Mod. Phys. 59 (1987), p. 639. [3] J. Miao et al, Nature 400 (1999), p. 342. [4] H. Lichte et al, Annu. Rev. Mater. Res. 37 (2007), p. 539. [5] J. Yamasaki et al, Appl. Phys. Lett. 101 (2012), 234105. [6] P Schniter and S Rangan. Signal Proc., IEEE Trans. on. 64(4), (2015), pp. 1043. [7] Supported by the Chemical Imaging, Signature Discovery, and Analytics in Motion initiatives at PNNL. PNNL is operated by Battelle Memorial Inst. for the US DOE; contract DE-AC05-76RL01830.

Schneider, Sebastian↗

Latent Twins

Over the past decade, scientific machine learning has transformed the development of mathematical and computational frameworks for analyzing, modeling, and predicting complex systems. From inverse problems to numerical partial differential equations (PDEs), dynamical systems, and model reduction, these advances have pushed the boundaries of what can be simulated. Yet they have often progressed in parallel, with representation learning and algorithmic solution methods evolving largely as separate pipelines. With Latent Twins, we propose a unifying mathematical framework that creates a hidden surrogate in latent space for the underlying equations. Whereas digital twins mirror physical systems in the digital world, Latent Twins mirror mathematical systems in a learned latent space governed by operators. Through this lens, classical modeling, inversion, model reduction, and operator approximation all emerge as special cases of a single principle. We establish the fundamental approximation properties of Latent Twins for both ordinary differential equations (ODEs) and PDEs and demonstrate the framework across three representative settings: (i) canonical ODEs, capturing diverse dynamical regimes; (ii) a PDE benchmark using the shallow-water equations, contrasting Latent Twin simulations with deep operator network and forecasts with a four-dimensional variational method baseline; and (iii) a challenging real-data geopotential reanalysis dataset, reconstructing and forecasting from sparse, noisy observations. Latent Twins provide a compact, interpretable surrogate for solution operators that evaluate across arbitrary time gaps in a single-shot, while remaining compatible with scientific pipelines such as assimilation, control, and uncertainty quantification. Looking forward, this framework offers scalable, theory-grounded surrogates that bridge data-driven representation learning and classical scientific modeling across disciplines.

Latent Twins↗

Interferometric SAR modelling of near surface data to improve geological model in the Surat Basin, Australia

Here we present a study on geophysical inverse modelling for subsurface structural properties of an unconventional hydrocarbon site that was monitored previously by Interferometric Synthetic Aperture Radar (InSAR) technology for surface deformation. A static three-dimensional geomodel along with extracted property maps replicates the depth of each underlying stratigraphic unit and structural feature with the density of each geological layer. We examine the hypothesis that integration of elastic properties of each formation layer with InSAR observations in a stratified elastic medium will lead to a viscoelastic geophysical inverse problem that can be solved to estimate fractional volume change at the reservoir level. Moreover, we examine synthetic scenarios in which the elastic properties of the formations are perturbed before determining the resulting impact on the rate of surface deformation. The results show that although the slope of underlying formations, their density and depth can define the extent and pattern of a deformation signal, their properties have a marginal impact on volumetric change compared to the dense network of shallow depth Coal Seam Gas (CSG) mining wells. Besides, it is also demonstrated that the inversion of InSAR deformation maps can resolve the uncertainties associated with low-resolution seismic interpretation as well as filling the data gaps within seismic acquisitions. A significant contribution of this investigation to the geological basin modelling involves a) introducing a remote and non-invasive technology such as InSAR to improve geophysical mapping of subsurface structures such as faults in areas with sparse or no reflective seismic information, and b) applying a multi-layer viscoelastic geophysical source model for an unconventional hydrocarbon reservoir such as CSG.

58 GEOSCIENCES↗

Deep Learning for Simultaneous Inference of Hydraulic and Transport Properties

Abstract Identification of a heterogeneous conductivity field and reconstruction of a contaminant release history are key aspects of subsurface remediation. These two goals are achieved by combining model predictions with sparse and noisy hydraulic head and concentration measurements. Solution of this inverse problem is notoriously difficult due to, in part, high dimensionality of the parameter space and high computational cost of repeated forward solves. We use a convolutional adversarial autoencoder (CAAE) to parameterize a heterogeneous non‐Gaussian conductivity field via a low‐dimensional latent representation. A three‐dimensional dense convolutional encoder‐decoder (DenseED) network serves as a forward surrogate of the flow and transport model. The CAAE‐DenseED surrogate is fed into the ensemble smoother with multiple data assimilation (ESMDA) algorithm to sample from the Bayesian posterior distribution of the unknown parameters, forming a CAAE‐DenseED‐ESMDA inversion framework. The resulting CAAE‐DenseED‐ESMDA inversion strategy is used to identify a three‐dimensional contaminant source and conductivity field. A comparison of the inversion results from CAAE‐ESMDA with physical flow and transport simulator and from CAAE‐DenseED‐ESMDA shows that the latter yields accurate reconstruction results at the fraction of the computational cost of the former.

Zhou, Zitong↗

Seismic Tomography 2024

Seismic tomography is the most abundant source of information about the internal structure of the Earth at scales ranging from a few meters to thousands of kilometers. It constrains the properties of active volcanoes, earthquake fault zones, deep reservoirs and storage sites, glaciers and ice sheets, or the entire globe. It contributes to outstanding societal problems related to natural hazards, resource exploration, underground storage, and many more. The recent advances in seismic tomography are being translated to nondestructive testing, medical ultrasound, and helioseismology. Nearly 50 yr after its first successful applications, this article offers a snapshot of modern seismic tomography. Focused on major challenges and particularly promising research directions, it is intended to guide both Earth science professionals and early-career scientists. The individual contributions by the coauthors provide diverse perspectives on topics that may at first seem disconnected but are closely tied together by a few coherent threads: multiparameter inversion for properties related to dynamic processes, data quality, and geographic coverage, uncertainty quantification that is useful for geologic interpretation, new formulations of tomographic inverse problems that address concrete geologic questions more directly, and the presentation and quantitative comparison of tomographic models. Finally, it remains to be seen which of these problems will be considered solved, solved to some extent, or practically unsolvable over the next decade.

58 GEOSCIENCES↗

Bayesian seismic inversion based on rock-physics prior modeling for the joint estimation of acoustic impedance, porosity and lithofacies

We propose a Bayesian approach for seismic inversion to estimate acoustic impedance, porosity and lithofacies within the reservoir conditioned to post-stack seismic and well data. The link between elastic and petrophysical properties is given by a joint prior distribution for the logarithm of impedance and porosity, based on a rock-physics model. The well conditioning is performed through a background model obtained by well log interpolation. Two different approaches are presented: in the first approach, the prior is defined by a single Gaussian distribution, whereas in the second approach it is defined by a Gaussian mixture to represent the well data multimodal distribution and link the Gaussian components to different geological lithofacies. The forward model is based on a linearized convolutional model. For the single Gaussian case, we obtain an analytical expression for the posterior distribution, resulting in a fast algorithm to compute the solution of the inverse problem, i.e. the posterior distribution of acoustic impedance and porosity as well as the facies probability given the observed data. For the Gaussian mixture prior, it is not possible to obtain the distributions analytically, hence we propose a Gibbs algorithm to perform the posterior sampling and obtain several reservoir model realizations, allowing an uncertainty analysis of the estimated properties and lithofacies. Both methodologies are applied to a real seismic dataset with three wells to obtain 3D models of acoustic impedance, porosity and lithofacies. The methodologies are validated through a blind well test and compared to a standard Bayesian inversion approach. Using the probability of the reservoir lithofacies, we also compute a 3D isosurface probability model of the main oil reservoir in the studied field.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Fossil Fuel Emission Verification Modeling at LLNL

We have an established project at LLNL to develop the tools needed to constrain fossil fuel carbon dioxide emissions using measurements of the carbon-14 isotope in atmospheric samples. In Figure 1 we show the fossil fuel plumes from Los Angeles and San Francisco for two different weather patterns. Obviously, a measurement made at any given location is going to depend on the weather leading up to the measurement. Thus, in order to determine the GHG emissions from some region using in situ measurements of those GHGs, we use state-of-the-art global and regional atmospheric chemistry-transport codes to simulate the plumes: the LLNL-IMPACT model (Rotman et al., 2004) and the WRFCHEM community code (http://www.wrf-model.org/index.php). Both codes can use observed (aka assimilated) meteorology in order to recreate the actual transport that occurred. The measured concentration of each tracer at a particular spatio-temporal location is a linear combination of the plumes from each region at that location (for non-reactive species). The challenge is to calculate the emission strengths for each region that fit the observed concentrations. In general this is difficult because there are errors in the measurements and modeling of the plumes. We solve this inversion problem using the strategy illustrated in Figure 2. The Bayesian Inference step combines the a priori estimates of the emissions, and their uncertainty, for each region with the results of the observations, and their uncertainty, and an ensemble of model predicted plumes for each region, and their uncertainty. The result is the mathematical best estimate of the emissions and their errors. In the case of non-linearities, or if we are using a statistical sampling technique such as a Markov Chain Monte Carlo technique, then the process is iterated until it converges (ie reaches stationarity). For the Bayesian inference we can use both a direct inversion capability, which is fast but requires assumptions of linearity and Gaussianity of errors, or one of several statistical sampling techniques, which are computationally slower but do not require either linearity or Gaussianity (Chow, et al., 2008; Delle Monache, et al., 2008). The emission regions we are using are based on the air-basins defined by the California Air Resources Board (CARB), see Figure 3. The only difference is that we have joined some of the smaller air basins together. The results of a test using 4 days of simulated observations using our ensemble retrieval system are shown in Figure 3 (right). The main source of the variation between the different model configurations arises from the uncertainty in the atmospheric boundary layer parameterization in the WRF model. We are currently developing a capability to constrain the boundary layer height in our carbon-14 work either by weighting the ensemble member results by the accuracy of their boundary layer height (using commercial aircraft observations), or as part of the retrieval process using an ensemble Kalman filter (EnKF) capability.

54 ENVIRONMENTAL SCIENCES↗