Search NASA⌕ Search

SEARCH · Search NASA

Results for “Inverse model”

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 613 records · Page 34

Statistical modelling and Bayesian inversion for a Compton imaging system: application to radioactive source localization

Abstract This paper presents a statistical forward model for a Compton imaging system, called Compton imager. This system, under development at the University of Illinois Urbana Champaign, is a variant of Compton cameras with a single type of sensors which can simultaneously act as scatterers and absorbers. This imager is convenient for imaging situations requiring a wide field of view. The proposed statistical forward model is then used to solve the inverse problem of estimating the location and energy of point-like sources from observed data. This inverse problem is formulated and solved in a Bayesian framework by using a Metropolis within Gibbs algorithm for the estimation of the location, and an expectation-maximization algorithm for the estimation of the energy. This approach leads to more accurate estimation when compared with the deterministic standard back-projection approach, with the additional benefit of uncertainty quantification in the low photon imaging setting.

Tarpau, Cécilia (ORCID:0000000286539490)↗

Characterizing the impact of model error in hydrologic time series recovery inverse problems

Hydrologic models are commonly over-smoothed relative to reality, owing to computational limitations and to the difficulty of obtaining accurate high-resolution information. When used in an inversion context, such models may introduce systematic biases which cannot be encapsulated by an unbiased “observation noise” term of the type assumed by standard regularization theory and typical Bayesian formulations. Despite its importance, model error is difficult to encapsulate systematically and is often neglected. In this paper, model error is considered for an important class of inverse problems that includes interpretation of hydraulic transients and contaminant source history inference: reconstruction of a time series that has been convolved against a transfer function (i.e., impulse response) that is only approximately known. Using established harmonic theory along with two results established here regarding triangular Toeplitz matrices, upper and lower error bounds are derived for the effect of systematic model error on time series recovery for both well-determined and over-determined inverse problems. It is seen that use of additional measurement locations does not improve expected performance in the face of model error. A Monte Carlo study of a realistic hydraulic reconstruction problem is presented, and the lower error bound is seen informative about expected behavior. Finally, a possible diagnostic criterion for blind transfer function characterization is also uncovered.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Regularization by denoising diffusion models for solving inverse PDE problems with application to full waveform inversion

Partial differential equation (PDE)-governed inverse problems are fundamental across various scientific and engineering applications; yet they face significant challenges due to nonlinearity, ill-posedness, and sensitivity to noise. Here, we introduce a computational framework, regularization by denoising using diffusion models for partial differential equations (RED-DiffEq), by integrating physics-driven inversion and data-driven learning. RED-DiffEq leverages pretrained diffusion models as a regularization mechanism for PDE-governed inverse problems. We apply RED-DiffEq to solve the full waveform inversion problem in geophysics, a challenging seismic imaging technique that seeks to reconstruct high-resolution subsurface velocity models from seismic measurement data. Our method shows enhanced accuracy and robustness compared to benchmark methods. Additionally, it exhibits strong generalization and domain decomposition capacity, enabling the inversion of more complex velocity models with larger domains than those used in training the diffusion model. Our framework can also be directly applied to diverse PDE-governed inverse problems.

Shan, Siming [Yale University, New Haven, CT (Unit↗

LAI inversion from optical reflectance using a neural network trained with a multiple scattering model

The inversion of the leaf area index (LAI) canopy parameter from optical spectral reflectance measurements is obtained using a backpropagation artificial neural network trained using input-output pairs generated by a multiple scattering reflectance model. The problem of LAI estimation over sparse canopies (LAI < 1.0) with varying soil reflectance backgrounds is particularly difficult. Standard multiple regression methods applied to canopies within a single homogeneous soil type yield good results but perform unacceptably when applied across soil boundaries, resulting in absolute percentage errors of >1000 percent for low LAI. Minimization methods applied to merit functions constructed from differences between measured reflectances and predicted reflectances using multiple-scattering models are unacceptably sensitive to a good initial guess for the desired parameter. In contrast, the neural network reported generally yields absolute percentage errors of <30 percent when weighting coefficients trained on one soil type were applied to predicted canopy reflectance at a different soil background.

Smith, James A.↗

Adjusting a Full Body Model to Mitigate Inverse Kinematics Artifacts in Opensim

BACKGROUND: In support of Vibration Isolation and Stabilization (VIS) system development for Human Health Countermeasures (HHC) exercise systems in space, such as the European Enhanced Exploration Exercise Device (E4D) [1], dynamic quantities required to model the response of a proposed VIS while considering the effect of VIS motion on the forces between the human and VIS platform were obtained using motion capture data [2]. On occasion, large-amplitude oscillatory spikes were found in the subject’s linear and angular momentum derivatives, affecting analyses that depend on forces and moments derived from motion capture. The purpose of this investigation was to identify causes of these artifacts and techniques for their resolution. METHODS AND RESULTS: To obtain the required human dynamic quantities to drive the VIS simulation, motion capture data was collected containing recorded trajectories of passive retroreflective markers on body landmarks of an exercising subject. Since the full body Rajagopal model [3] was originally used to enhance gait analysis, upper body joints did not require large Ranges of Motion (ROM). We thus modified the Rajagopal model [4, 5] to allow it to be used for upper body intensive exercises like those common to the E4D. OpenSim Inverse Kinematics (IK) [6] was performed using these scaled subject models to generate the joint angles throughout the exercise while minimizing marker error. At times, the arms were observed to ‘snap’ from one configuration to another, causing spike artifacts. Following IK, a custom OpenSim plugin [7] was used to determine the required dynamic quantities including the rates of change of the linear and angular momenta of the human. Since motion capture is recorded at a larger time step than required by the VIS simulation, the human center of mass location was fit with splines and a second derivative taken to obtain the momentum derivative, allowing the VIS simulation to maintain conservation of momentum when appropriate. In a few cases during this stage, artifacts much larger than the expected noise of the second derivatives were introduced. Investigation of cases containing artifacts revealed several modifications that could be made to the OpenSim model to improve IK results. Since OpenSim models use Euler angles and rotation sequences, ‘gimbal lock’ would be encountered in the arms when raised 90 degrees to the side (e.g., T-pose, some hang clean exercise, etc.). This was resolved by reorienting the horizontal axes at the shoulder joint by 45 degrees, placing ‘gimbal lock’ outside common arm ROM, with the arm ROMs adjusted following this change. Elbow and wrist ROMs could also be adjusted to allow realistic motion while at the same time limiting the likelihood of unrealistic orientations. On occasion, the arms flipped backwards when raised above the head. This was prevented by using medial elbow markers in scaling and IK. When medial markers were not available, the acromial joint location in the unscaled model was shifted before model scaling to better align the arm with available markers. Lastly, artifacts which became apparent after taking the second derivatives of spline-fit data were found to occur in cases when the pelvis rotation limit prevented the full range of motion of an exercise. These issues were resolved by unclamping the pelvis rotation limit. Through this investigation, an understanding of conditions leading to IK artifacts was acquired allowing the automation of artifact detection. These artifact detection and mitigation techniques can be applied toward modeling of upper body motions in aerospace and other fields for improved IK results.

C A Bell↗

INVERSE COMPTON SCATTERING MODEL FOR X-RAY EMISSION OF THE GAMMA-RAY BINARY LS 5039

We propose a model for the gamma-ray binary LS 5039 in which the X-ray emission is due to the inverse Compton (IC) process instead of the synchrotron radiation. Although the synchrotron model has been discussed in previous studies, it requires a strong magnetic field which leads to a severe suppression of the TeV gamma-ray flux in conflict with H.E.S.S. observations. In this paper, we calculate the IC emission by low energy electrons ({gamma}{sub e} {approx}< 10{sup 3}) in the Thomson regime. We find that IC emission of the low energy electrons can explain the X-ray flux and spectrum observed with Suzaku if the minimum Lorentz factor of injected electrons {gamma}{sub min} is around 10{sup 3}. In addition, we show that the Suzaku light curve is well reproduced if {gamma}{sub min} varies in proportion to the Fermi flux when the distribution function of injected electrons at higher energies is fixed. We conclude that the emission from LS 5039 is well explained by the model with the IC emission from electrons whose injection properties are dependent on the orbital phase. Since the X-ray flux is primarily determined by the total number of cooling electrons, this conclusion is rather robust, although some mismatches between the model and observations at the GeV band remain in the present formulation.

79 ASTRONOMY AND ASTROPHYSICS↗

A model for the temperature inversion within the atmosphere of Saturn

A model for the temperature inversion within the atmosphere of Saturn is proposed and is shown to be consistent with photometric data in the 17- to 25-micron region. The proposed model incorporates solar heating by some 'aerosol', with the aerosol heating per unit mass of the atmosphere being uniformly distributed throughout that portion of the atmosphere overlying the upper cloud deck. For a methane-to-hydrogen mixing ratio of 0.0007, the model results suggest that 20% of the incident solar radiation is absorbed by the aerosol, while this is reduced to 16% for an enhanced methane mixing ratio of 0.0021.

Tokunaga, A.↗

Analytic model approach to the inversion of scattering data

An analytic model approach is applied to several simple atmospheric inversion problems. This method gives a sharp determination of aerosol size distribution parameters. It is shown that this analytic approach, together with ground level point sampling data measurements, can be used to infer information on the tropospheric ozone profile.

Green, A. E. S.↗

RLMolLM: Reinforcement Learning-Enhanced Language Model Framework for Inverse Molecular Design

Inverse molecular design faces significant challenges due to vast chemical space and complex property requirements. While language models show promise for molecular generation, they struggle with validity, multi-property optimization, and structural constraints. This work presents RLMolLM, a reinforcement learning framework combining Proximal Policy Optimization (PPO) with genetic algorithms to address these limitations. Our approach optimizes multiple user-specified properties including quantitative estimates of drug-likeness (QED), synthetic accessibility (SA), and ADMET (absorption, distribution, metabolism, excretion, and toxicity) endpoints without requiring complete model retraining, while maintaining capability for scaffold-constrained generation where specific substructures must be preserved. We outperform state-of-the-art methods for molecular optimization, achieving best QED scores across GDB13, Moses, and Zinc datasets with up to 31% improvement over previous methods while maintaining excellent validity, uniqueness, and novelty metrics. For simultaneous multi-property optimization, our framework achieves substantial improvements in ADMET properties including 4.5-fold reduction in hERG toxicity and enhanced Caco-2 permeability compared to Moses dataset. Under structural constraints, the framework significantly improves molecular validity while preserving scaffolds and effectively optimizing properties. In conclusion, this versatile solution advances pharmaceutical and materials molecular design through effective integration of reinforcement learning and genetic algorithms with multi-property optimization and scaffold preservation.

Genetic algorithms↗

A framework for data-driven solution and parameter estimation of PDEs using conditional generative adversarial networks

We employ and adapt the image-to-image translation concept based on conditional generative adversarial networks (cGAN) for learning a forward and an inverse solution operator of partial differential equations (PDEs). We focus on steady-state solutions of coupled hydromechanical processes in heterogeneous porous media and present the parameterization of the spatially heterogeneous coefficients, which is exceedingly difficult using standard reduced-order modeling techniques. We show that our framework provides a speed-up of at least 2,000 times compared to a finite-element solver and achieves a relative root-mean-square error (r.m.s.e.) of less than 2% for forward modeling. For inverse modeling, the framework estimates the heterogeneous coefficients, given an input of pressure and/or displacement fields, with a relative r.m.s.e. of less than 7%, even for cases where the input data are incomplete and contaminated by noise. The framework also provides a speed-up of 120,000 times compared to a Gaussian prior-based inverse modeling approach while also delivering more accurate results.

97 MATHEMATICS AND COMPUTING↗

Seismic Modeling Workshop - Joint inversion [Slides]

Why joint inversion? It enables spanning spatial and data scales! Multiple benefits: Different data sets have different spatial coverage and resolution. “Standard” geophysical models are developed only to fit one type of data. Different data types have different strengths. Multiple challenges: Deal with different data bandwidths. Design responsive misfit norms; relative weighting of data sets. Make assumptions to model the different data; relationships between independent data sets.

58 GEOSCIENCES↗

Computation of Jupiter interior models from gravitational inversion theory

Spacecraft measurements of Jupiter have provided the mass, standard pressure level radius, rotation law, internal mass distribution multipole moments, and internal composition and temperature distribution constraints, for the present implementation of a method for deriving planetary interior models that exactly satisfy a set of N gravitational constraints by means of appropriate iteration. The models are not forced to fit the more indirectly derived constraints, which are instead used as conistency checks. In the case of an He mass fraction in the envelope Y of 0.2, the inferred pressure at a mass density of about 0.2 g/cu cm is about a factor of 2 higher than would be indicated by experimental H compression data in the relevant pressure range of 100,000 to one million bar. The inferred pressure distribution is in better agreement with the shock data for a nominal Y value of 0.3 + or - 0.05.

Hubbard, W. B.↗

An empirical model and an inversion technique for radar scattering from bare soil surfaces

Polarimetric radar measurements were conducted for bare soil surfaces under a variety of roughness and moisture conditions at L-, C-, and X-band frequencies at incidence angles ranging from 10 to 70 deg. Using a laser profiler and dielectric probes, a complete and accurate set of ground truth data were collected for each surface condition, from which accurate measurements were made of the rms height, correlation length, and dielectric constant. Based on knowledge of the scattering behavior in limiting cases and the experimental observations, an empirical model was developed which was found to yield very good agreement with the backscattering measurements of this study, as well as with measurements reported in other investigations. An inversion technique for predicting the rms height of the surface and its moisture content from multipolarized radar observations is developed on the basis of the model.

Oh, Yisok↗

Circuit model for the inverse Z-pinch wire array switch.

A 0D circuit code is introduced to study the wire array switch concept introduced in. It has been implemented and researched at Imperial College. An exploding wire array, the switch, is in parallel with the load, an imploding wire array. Most of the current flows in the exploding array until it expands and becomes highly resistive. The 0D code contains simple models of Joule energy deposition and plasma expansion for W and Al wires. The purpose of the device is to produce fast Z-pinch implosion, below 100ns on MAGPIE and the Sandia Z machine. Self and mutual inductances are taken into consideration as well as the rocket model for wire ablation. The switch characteristics of the exploding array are prescribed and tuned up to agree with MAGPIE shots. The dependence of the device on the configuration of the arrays is studied and scaling to ZR conditions is explored.

33 ADVANCED PROPULSION SYSTEMS↗

The effect of error models in the multiscale inversion of binary permeability fields.

We present results from a recently developed multiscale inversion technique for binary media, with emphasis on the effect of subgrid model errors on the inversion. Binary media are a useful fine-scale representation of heterogeneous porous media. Averaged properties of the binary field representations can be used to characterize flow through the porous medium at the macroscale. Both direct measurements of the averaged properties and upscaling are complicated and may not provide accurate results. However, it may be possible to infer upscaled properties of the binary medium from indirect measurements at the coarse scale. Multiscale inversion, performed with a subgrid model to connect disparate scales together, can also yield information on the fine-scale properties. We model the binary medium using truncated Gaussian fields, and develop a subgrid model for the upscaled permeability based on excursion sets of those fields. The subgrid model requires an estimate of the proportion of inclusions at the block scale as well as some geometrical parameters of the inclusions as inputs, and predicts the effective permeability. The inclusion proportion is assumed to be spatially varying, modeled using Gaussian processes and represented using a truncated Karhunen-Louve (KL) expansion. This expansion is used, along with the subgrid model, to pose as a Bayesian inverse problem for the KL weights and the geometrical parameters of the inclusions. The model error is represented in two different ways: (1) as a homoscedastic error and (2) as a heteroscedastic error, dependent on inclusion proportionality and geometry. The error models impact the form of the likelihood function in the expression for the posterior density of the objects of inference. The problem is solved using an adaptive Markov Chain Monte Carlo method, and joint posterior distributions are developed for the KL weights and inclusion geometry. Effective permeabilities and tracer breakthrough times at a few 'sensor' locations (obtained by simulating a pump test) form the observables used in the inversion. The inferred quantities can be used to generate an ensemble of permeability fields, both upscaled and fine-scale, which are consistent with the observations. We compare the inferences developed using the two error models, in terms of the KL weights and fine-scale realizations that could be supported by the coarse-scale inferences. Permeability differences are observed mainly in regions where the inclusions proportion is near the percolation threshold, and the subgrid model incurs its largest approximation. These differences also reflected in the tracer breakthrough times and the geometry of flow streamlines, as obtained from a permeameter simulation. The uncertainty due to subgrid model error is also compared to the uncertainty in the inversion due to incomplete data.

58 GEOSCIENCES↗

GenAI4UQ: A software for forward and inverse uncertainty quantification using conditional generative AI

We introduce GenAI4UQ, a software package for forward and inverse uncertainty quantification in model calibration, parameter estimation, and ensemble forecasting. GenAI4UQ leverages a generative AI-based conditional modeling framework to address limitations of traditional inverse modeling techniques, such as Markov Chain Monte Carlo (MCMC) methods. By replacing computationally intensive iterative processes with a direct, learned mapping, GenAI4UQ enables efficient calibration of input parameters and generation of predictions directly from observations. The software supports rapid ensemble forecasting with robust uncertainty quantification while maintaining computational and storage efficiency. Built-in auto-tuning of hyperparameters simplifies model training, ensuring accessibility for users with varying expertise. Its versatile conditional generative framework is applicable across diverse scientific domains. While GenAI4UQ offers significant advantages in flexibility and efficiency, users should interpret its uncertainty estimates with caution in data-sparse scenarios, as the model may overestimate uncertainty—an effect common to all surrogate-based approaches including MCMC with surrogate models. Despite this, GenAI4UQ transforms inverse modeling by providing a fast, reliable, and user-friendly solution. It empowers researchers and practitioners to quickly estimate parameter distributions and generate model predictions for new observations, facilitating efficient decision-making and advancing the state of uncertainty quantification in computational modeling.

97 MATHEMATICS AND COMPUTING↗

Solving inverse problems in stochastic models using deep neural networks and adversarial training

Inverse problems associated with stochastic models constitute a significant portion of scientific and engineering applications. In such cases the unknown quantities are distributions. The applicability of traditional methods is limited because of their demanding assumptions or prohibitive computational consumption; for example, maximum likelihood methods require closed-form density functions, and Markov Chain Monte Carlo needs a large number of simulations. We propose a new method that estimates the unknown distribution by matching the statistical properties between observed and simulated random processes. We leverage the expressive power of neural networks to approximate the unknown distribution and use a discriminative neural network for computing the statistical discrepancies between the observed and simulated random processes. Here we demonstrated numerically that the proposed methods can estimate both the model parameters and learn complicated unknown distributions.

42 ENGINEERING↗