Best Predictive GLMM using LASSO with Application on High-Speed Network
Explore the source record for details and available documents.
SEARCH · Search NASA
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.
Explore the source record for details and available documents.
This study demonstrates a robust trajectory optimization approach for rendezvous and proximity operations with angles-only navigation measurements. Often, sensors that directly measure relative range and velocity require communication or coordination between the chaser and target vehicle and can have limiting pointing accuracy, mass, or power requirements compared to angle measurement sensors. Thus, the capability to perform a rendezvous with only angle measurements can be advantageous for vehicle design and to improve robustness to failures. However, the well studied limitation of angles-only navigation in measuring range results in large uncertainties in the navigation system that must be reduced with chaser vehicle thrust maneuvers to induce observability in range for the navigation filter. This analysis presents a trajectory optimization problem for a lunar ascent rendezvous during a crewed lunar mission in a Near-Rectilinear Halo Orbit (NRHO) that is limited to only angle measurements. The objective of this study is to show that an angles-only rendezvous is feasible in an NRHO and to present the sensitivity to an assortment of constraints generated from a systematic optimization process using linear covariance analysis and particle swarm optimization. Linearized NRHO dynamics and linearized relative targeting are applied to use linear covariance analysis to determine the expected delta-v and trajectory dispersions due to initial state uncertainty, sensor errors, maneuver execution errors, and unmodeled dynamics. The delta-v and trajectory dispersions are passed into a particle swarm optimization algorithm to find the optimized maneuver profile that minimizes fuel use while satisfying constraints such as free drift and underburn to 3-sigma certainty. The trajectory constraints including time available, desired final uncertainty, and initial uncertainty are varied to ascertain sensitivity and desirable engineering trades.
Recent literature in applied estimation theory reflects growing interest in the sigma-point (also called unscented ) formulation for optimal sequential state estimation, often describing performance comparisons with extended Kalman filters as applied to specific dynamical problems [c.f. 1, 2, 3]. Favorable attributes of sigma-point filters are described as including a lower expected error for nonlinear even non-differentiable dynamical systems, and a straightforward formulation not requiring derivation or implementation of any partial derivative Jacobian matrices. These attributes are particularly attractive, e.g. in terms of enabling simplified code architecture and streamlined testing, in the formulation of estimators for nonlinear spaceflight mechanics systems, such as filter software onboard deep-space robotic spacecraft. As presented in [4], the Sigma-Point Consider Filter (SPCF) algorithm extends the sigma-point filter algorithm to the problem of consider covariance analysis. Considering parameters in a dynamical system, while estimating its state, provides an upper bound on the estimated state covariance, which is viewed as a conservative approach to designing estimators for problems of general guidance, navigation and control. This is because, whether a parameter in the system model is observable or not, error in the knowledge of the value of a non-estimated parameter will increase the actual uncertainty of the estimated state of the system beyond the level formally indicated by the covariance of an estimator that neglects errors or uncertainty in that parameter. The equations for SPCF covariance evolution are obtained in a fashion similar to the derivation approach taken with standard (i.e. linearized or extended) consider parameterized Kalman filters (c.f. [5]). While in [4] the SPCF and linear-theory consider filter (LTCF) were applied to an illustrative linear dynamics/linear measurement problem, in the present work examines the SPCF as applied to nonlinear sequential consider covariance analysis, i.e. in the presence of nonlinear dynamics and nonlinear measurements. A simple SPCF for orbit determination, exemplifying an algorithm hosted in the guidance, navigation and control (GN&C) computer processor of a hypothetical robotic spacecraft, was implemented, and compared with an identically-parameterized (standard) extended, consider-parameterized Kalman filter. The onboard filtering scenario examined is a hypothetical spacecraft orbit about a small natural body with imperfectly-known mass. The formulations, relative complexities, and performances of the filters are compared and discussed.
We define a new indirect measure of the diffusion model error called the diffusion model error source. When this model error source is added to the diffusion equation, the transport solution for the angular-integrated intensity is obtained. This source represents a means by which a transport code can be used to generate information relating to the adequacy of diffusion theory for any given problem without actually solving the diffusion equation. The generation of this source does not relate in any way to acceleration of the iterative convergence of transport solutions. Perhaps the most well-known indirect measure of the diffusion model error is the variable-Eddington tensor. This tensor provides a great deal of information about the angular dependence of the angular intensity solution, but it is not always simple to interpret. In contrast, our diffusion model error source is a scalar that is conceptually easy to understand. In addition to defining the diffusion model error source analytically, we show how to generate this source numerically relative to the S{sub n} radiative transfer equations with linear-discontinuous spatial discretization. This numerical source is computationally tested and shown to reproduce the Sn solution for a Marshak-wave problem. (authors)
We propose a scheme for entangling two photons via the quantum Zeno effect, which describes the inhibition of quantum evolution by frequent measurements and is based on the difference between summing amplitudes and probabilities. For a given error probability P{sub error}, our scheme requires that the one-photon loss rate {xi}{sub 1{gamma}} and the two-photon absorption rate {xi}{sub 2{gamma}} in some medium satisfy {xi}{sub 1{gamma}}/{xi}{sub 2{gamma}}=2P{sub error}{sup 2}/{pi}{sup 2}, which is significantly improved compared to previous approaches. Again based on the quantum Zeno effect, as well as coherent excitations, we present a possibility to fulfill this requirement in an otherwise linear optics setup.
We present a method for improving the tolerance of a conventional LQG controller to parameter errors in the plant model. The improvement is achieved by introducing additional terms reflecting the structure of the parameter errors into the LQR cost function, and also the process and measurement noise models. Adjusting the sizes of these additional terms permits a trade-off between robustness and nominal performance. Manipulation of some of the additional terms leads to high gain controllers while other terms lead to low gain controllers. Conditions are developed under which the high-gain approach asymptotically recovers the robustness of the corresponding full-state feedback design, and the low-gain approach makes the closed-loop poles asymptotically insensitive to parameter errors.
In a coded communication system with equiprobable signaling, MLD minimizes the word error probability and delivers the most likely codeword associated with the corresponding received sequence. This decoding has two drawbacks. First, minimization of the word error probability is not equivalent to minimization of the bit error probability. Therefore, MLD becomes suboptimum with respect to the bit error probability. Second, MLD delivers a hard-decision estimate of the received sequence, so that information is lost between the input and output of the ML decoder. This information is important in coded schemes where the decoded sequence is further processed, such as concatenated coding schemes, multi-stage and iterative decoding schemes. In this chapter, we first present a decoding algorithm which both minimizes bit error probability, and provides the corresponding soft information at the output of the decoder. This algorithm is referred to as the MAP (maximum aposteriori probability) decoding algorithm.
Purpose: The objective of this study is to establish the feasibility of using radiation-induced acoustics to measure the range and Bragg peak dose from a pulsed proton beam. Simulation studies implementing a prototype scanner design based on computed tomographic methods were performed to investigate the sensitivity to proton range and integral dose. Methods: Derived from thermodynamic wave equation, the pressure signals generated from the dose deposited from a pulsed proton beam with a 1 cm lateral beam width and a range of 16, 20, and 27 cm in water using Monte Carlo methods were simulated. The resulting dosimetric images were reconstructed implementing a 3D filtered backprojection algorithm and the pressure signals acquired from a 71-transducer array with a cylindrical geometry (30 × 40 cm) rotated over 2π about its central axis. Dependencies on the detector bandwidth and proton beam pulse width were performed, after which, different noise levels were added to the detector signals (using 1 μs pulse width and a 0.5 MHz cutoff frequency/hydrophone) to investigate the statistical and systematic errors in the proton range (at 20 cm) and Bragg peak dose (of 1 cGy). Results: The reconstructed radioacoustic computed tomographic image intensity was shown to be linearly correlated to the dose within the Bragg peak. And, based on noise dependent studies, a detector sensitivity of 38 mPa was necessary to determine the proton range to within 1.0 mm (full-width at half-maximum) (systematic error < 150 μm) for a 1 cGy Bragg peak dose, where the integral dose within the Bragg peak was measured to within 2%. For existing hydrophone detector sensitivities, a Bragg peak dose of 1.6 cGy is possible. Conclusions: This study demonstrates that computed tomographic scanner based on ionizing radiation-induced acoustics can be used to verify dose distribution and proton range with centi-Gray sensitivity. Realizing this technology into the clinic has the potential to significantly impact beam commissioning, treatment verification during particle beam therapy and image guided techniques.
We consider Linear Programming (LP) solution of a Compressed Sensing (CS) problem over reals, also known as the Basis Pursuit (BasP) algorithm. The BasP allows interpretation as a channel-coding problem, and it guarantees the error-free reconstruction over reals for properly chosen measurement matrix and sufficiently sparse error vectors. In this manuscript, we examine how the BasP performs on a given measurement matrix and develop a technique to discover sparse vectors for which the BasP fails. The resulting algorithm is a generalization of our previous results on finding the most probable error-patterns, so called instantons, degrading performance of a finite size Low-Density Parity-Check (LDPC) code in the error-floor regime. The BasP fails when its output is different from the actual error-pattern. We design CS-Instanton Search Algorithm (ISA) generating a sparse vector, called CS-instanton, such that the BasP fails on the instanton, while its action on any modification of the CS-instanton decreasing a properly defined norm is successful. We also prove that, given a sufficiently dense random input for the error-vector, the CS-ISA converges to an instanton in a small finite number of steps. Performance of the CS-ISA is tested on example of a randomly generated 512 * 120 matrix, that outputs the shortest instanton (error vector) pattern of length 11.
Isosurface visualization is fundamental for exploring and analyzing 3D volumetric data. Marching cubes (MC) algorithms with linear interpolation are commonly used for isosurface extraction and visualization. Although linear interpolation is easy to implement, it has limitations when the underlying data is complex and high-order, which is the case for most real-world data. Linear interpolation can output vertices at the wrong location. Its inability to deal with sharp features and features smaller than grid cells can lead to an incorrect isosurface with holes and broken pieces. Despite these limitations, isosurface visualizations typically do not include insight into the spatial location and the magnitude of these errors. We utilize high-order interpolation methods with MC algorithms and interactive visualization to highlight these uncertainties. Our visualization tool helps identify the regions of high interpolation errors. It also allows users to query local areas for details and compare the differences between isosurfaces from different interpolation methods. In addition, we employ high-order methods to identify and reconstruct possible features that linear methods cannot detect. We showcase how our visualization tool helps explore and understand the extracted isosurface errors through synthetic and real-world data.
In contrast to the classic view of static DNA double-strand breaks (DSBs) being repaired at the site of damage, we hypothesize that DSBs move and merge with each other over large distances (m). As X-ray dose increases, the probability of having DSB clusters increases as does the probability of misrepair and cell death. Experimental work characterizing the X-ray dose dependence of radiation-induced foci (RIF) in nonmalignant human mammary epithelial cells (MCF10A) is used here to validate a DSB clustering model. We then use the principles of the local effect model (LEM) to predict the yield of DSBs at the submicron level. Two mechanisms for DSB clustering, namely random coalescence of DSBs versus active movement of DSBs into repair domains are compared and tested. Simulations that best predicted both RIF dose dependence and cell survival after X-ray irradiation favored the repair domain hypothesis, suggesting the nucleus is divided into an array of regularly spaced repair domains of ~;;1.55 m sides. Applying the same approach to high-linear energy transfer (LET) ion tracks, we are able to predict experimental RIF/m along tracks with an overall relative error of 12percent, for LET ranging between 30 350 keV/m and for three different ions. Finally, cell death was predicted by assuming an exponential dependence on the total number of DSBs and of all possible combinations of paired DSBs within each simulated RIF. Relative biological effectiveness (RBE) predictions for cell survival of MCF10A exposed to high-LET showed an LET dependence that matches previous experimental results for similar cell types. Overall, this work suggests that microdosimetric properties of ion tracks at the submicron level are sufficient to explain both RIF data and survival curves for any LET, similarly to the LEM assumption. Conversely, high-LET death mechanism does not have to infer linear-quadratic dose formalism as done in the LEM. In addition, the size of repair domains derived in our model are based on experimental RIF and are three times larger than the hypothetical LEM voxel used to fit survival curves. Our model is therefore an alternative to previous approaches that provides a testable biological mechanism (i.e., RIF). In addition, we propose that DSB pairing will help develop more accurate alternatives to the linear cancer risk model (LNT) currently used for regulating exposure to very low levels of ionizing radiation.
One of the key components of a Fourier Transform Infrared Spectrometer (FTIR) is the linear translation stage used to vary the optical path length between the two arms of the interferometer. This translation mechanism must produce extremely constant velocity motion across its entire range of travel to allow the instrument to attain high signal-to-noise ratio and spectral resolving power. A new spectrometer is being developed at the Jet Propulsion Laboratory under NASA s Planetary Instrument Definition and Development Program (PIDDP). The goal of this project is to build upon existing spaceborne FTIR spectrometer technology to produce a new instrument prototype that has drastically superior spectral resolution and substantially lower mass, making it feasible for planetary exploration. In order to achieve these goals, Alliance Spacesystems, Inc. (ASI) has developed a linear translation mechanism using a novel ultrasonic piezo linear motor in conjunction with a fully kinematic, fault tolerant linear rail system. The piezo motor provides extremely smooth motion, is inherently redundant, and is capable of producing unlimited travel. The kinematic rail uses spherical Vespel(R). rollers and bushings, which eliminates the need for wet lubrication, while providing a fault tolerant platform for smooth linear motion that will not bind under misalignment or structural deformation. This system can produce velocities from 10 - 100 mm/s with less than 1% velocity error over the entire 100-mm length of travel for a total mechanism mass of less than 850 grams. This system has performed over half a million strokes under vacuum without excessive wear or degradation in performance. This paper covers the design, development, and testing of this linear translation mechanism as part of the Planetary Atmosphere Occultation Spectrometer (PAOS) instrument prototype development program.
The proposed size-independent linear-elastic fracture toughness, K (sub Isi), for potential inclusion in ASTM E399 targets a consistent 0.5 millimeters crack extension for all specimen sizes through an offset secant that is a function of the specimen ligament length. The K (sub Isi) method also includes an increase in allowable deformation, and the removal of the P (sub max)/P (sub Q) criterion. A finite element study of the K (sub Isi) test method confirms the viability of the increased deformation limit, but has also revealed a few areas of concern. Findings: 1. The deformation limit, b (sub o) greater than or equal to 1.1 times (K (sub I) divided by delta (sub ys) squared) maintains a K-dominant crack tip field with limited plastic contribution to the fracture energy; 2. The three dimensional effects on compliance and the shape of the force versus CMOD (Crack-Mouth Opening Displacement) trace are significant compared to a plane strain assumption; 3. The non-linearity in the force versus CMOD trace at deformations higher than the current limit of 2.5 times (K (sub I) divided by delta (sub ys) squared) is sufficient to introduce error or even "false calls" regarding crack extension when using a constant offset secant line. This issue is more significant for specimens with W (width) greater than or equal to 2 inches; 4. A non-linear plasticity correction factor in the offset secant may improve the viability of the method at deformations between 2.5 times (K (sub I) divided by delta (sub ys) squared) and 1.1 times (K (sub I) divided by delta (sub ys) squared).
Systems using inertia wheels are evaluated in this report to determine their suitability for precise attitude control of a satellite and to select superior system configurations. Various possible inertia wheel system configurations are first discussed in a general manner. Three of these systems which appear more promising than the others are analyzed in detail, using the Orbiting Astronomical Observatory as an example. The three systems differ from each other only by the method of damping, which is provided by either a rate gyro, an error-rate network, or a tachometer in series with a high-pass filter. An analytical investigation which consists of a generalized linear analysis, a nonlinear analysis using the switching-time method, and an analog computer study shows that all three systems are theoretically capable of producing adequate response and also of maintaining the required pointing accuracy for the Orbiting Astronomical Observatory of plus or minus 0.1 second of arc. Practical considerations and an experimental investigation show, however, that the system which uses an error-rate network to provide damping is superior to the other two systems. The system which uses a rate gyro is shown to be inferior because the threshold level causes a significant amount of limit-cycle operation, and the system which uses a tachometer with a filter is shown to be inferior because a device with the required dynamic range of operation does not appear to be available. The experimental laboratory apparatus used to investigate the dynamic performance of the systems is described, and experimental results are included to show that under laboratory conditions with relatively large extraneous disturbances, a dynamic tracking error of less than plus or minus 0.5 second of arc was obtained.
Partial least squares regression (PLSR) and support vector regression (SVR) models were optimized for the quantification of U(VI) (10–320 g L −1 ) and HNO 3 (0.6–6 M) by Raman spectroscopy with optimized calibration sets chosen by optimal design of experiments. The designed approach effectively minimized the number of samples in the calibration set for PLSR and SVR by selecting sample concentrations with a quadratic process model, despite complex confounding and covarying spectral features in the spectra. The top PLS2 model resulted in percent root mean square errors of prediction for U(VI), HNO 3 , and NO 3 − of 3.7%, 3.6%, and 2.9%, respectively. PLS1 models performed similarly despite modeling an analyte with a majority linear response (i.e., uranyl symmetric stretch) and another with more covarying vibrational modes (i.e., HNO 3 ). Partial least squares (PLS) model loadings and regression coefficients were evaluated to better understand the relationship between weaker Raman bands and covarying spectral features. Support vector machine models outperformed PLS1 models, resulting in percent root mean square error of prediction values for U(VI) and HNO 3 of 1.5% and 3.1%, respectively. The optimal nonlinear SVR model was trained using a similar number of samples (11) compared with the PLSR model, even though PLS is a linear modeling approach. The generic D-optimal design presented in this work provides a robust statistical framework for selecting training set samples in disparate two-factor systems. This approach reinforces Raman spectroscopy for the quantification of species relevant to the nuclear fuel cycle and provides a robust chemometric modeling approach to bolster online monitoring in challenging process environments.
Particulate matter air pollution is a leading cause of global mortality, particularly in Asia and Africa. Addressing the high and wide-ranging air pollution levels requires ambient monitoring, but many low- and middle-income countries (LMICs) remain scarcely monitored. To address these data gaps, recent studies have utilized low-cost sensors. These sensors have varied performance, and little literature exists about sensor intercomparison in Africa. By colocating 2 QuantAQ Modulair-PM, 2 PurpleAir PA-II SD, and 16 Clarity Node-S Generation II monitors with a reference-grade Teledyne monitor in Accra, Ghana, we present the first intercomparisons of different brands of low-cost sensors in Africa, demonstrating that each type of low-cost sensor PM2.5 is strongly correlated with reference PM2.5, but biased high for ambient mixture of sources found in Accra. When compared to a reference monitor, the QuantAQ Modulair-PM has the lowest mean absolute error at 3.04 μg/m3, followed by PurpleAir PA-II (4.54 μg/m3) and Clarity Node-S (13.68 μg/m3). We also compare the usage of 4 statistical or machine learning models (Multiple Linear Regression, Random Forest, Gaussian Mixture Regression, and XGBoost) to correct low-cost sensors data, and find that XGBoost performs the best in testing (R2: 0.97, 0.94, 0.96; mean absolute error: 0.56, 0.80, and 0.68 μg/m3 for PurpleAir PA-II, Clarity Node-S, and Modulair-PM, respectively), but tree-based models do not perform well when correcting data outside the range of the colocation training. Therefore, we used Gaussian Mixture Regression to correct data from the network of 17 Clarity Node-S monitors deployed around Accra, Ghana, from 2018 to 2021. We find that the network daily average PM2.5 concentration in Accra is 23.4 μg/m3, which is 1.6 times the World Health Organization Daily PM2.5 guideline of 15 μg/m3. While this level is lower than those seen in some larger African cities (such as Kinshasa, Democratic Republic of the Congo), mitigation strategies should be developed soon to prevent further impairment to air quality as Accra, and Ghana as a whole, rapidly grow.
Purpose: To investigate the accuracy and potential limitations of MV image-based dynamic multileaf collimator (DMLC) tracking in a porcine model on a linear accelerator. Methods and Materials: A thermo-expandable NiTi stent designed for kilovoltage (kV) X-ray visualization of lung lesions was inserted into the bronchia of three anaesthetized Goettingen minipigs. A four-dimensional computed tomography scan was used for planning a five-field conformal treatment with circular multileaf collimator (MLC) apertures. A 22.5 Gy single fraction treatment was delivered to the pigs. The peak-to-peak stent motion was 3 to 8 mm, with breathing periods of 1.2 to 4 s. Before treatment, X-ray images were used for image-guided setup based on the stent. During treatment delivery, continuous megavoltage (MV) portal images were acquired at 7.5 Hz. The stent was segmented in the images and used for continuous adaptation of the MLC aperture. Offline, the tracking error in beam's eye view of the treatment beam was calculated for each MV image as the difference between the MLC aperture center and the segmented stent position. The standard deviations of the systematic error {Sigma} and the random error {sigma} were determined and compared with the would-be errors for a nontracking treatment with pretreatment image-guided setup. Results: Reliable stent segmentation was obtained for 11 of 15 fields. Segmentation failures occurred when image contrast was dominated by overlapping anatomical structures (ribs, diaphragm) rather than by the stent, which was designed for kV rather than MV X-ray visibility. For the 11 fields with reliable segmentation, {Sigma} was 0.5 mm/0.4 mm in the two imager directions, whereas {sigma} was 0.5 mm/1.1 mm. Without tracking, {Sigma} and {sigma} would have been 1.7 mm/1.4 mm and 0.8 mm/1.4 mm, respectively. Conclusion: For the first time, in vivo DMLC tracking has been demonstrated on a linear accelerator showing the potential for improved targeting accuracy. The study mimicked the envisioned patient workflow of future patient treatments. Clinical implementation of MV image-based tracking would require markers designed for MV visibility.
In numerous applications, surrogate models are used as a replacement for accurate parameter-to-observable mappings when solving large-scale inverse problems governed by partial differential equations (PDEs). The surrogate model may be a computationally cheaper alternative to the accurate parameter-to-observable mappings and/or may ignore additional unknowns or sources of uncertainty. The Bayesian approximation error (BAE) approach provides a means to account for the induced uncertainties and approximation errors, i.e. the errors between the accurate parameter-to-observable mapping and the surrogate. The statistics of these errors are, however, in general unknown a priori, and are thus calculated using Monte Carlo sampling. Although the sampling is typically carried out offline, i.e. before considering the data, the process can still represent a computational bottleneck. In this work, we develop a scalable computational approach for reducing the costs associated with the sampling stage of the BAE approach. Specifically, we consider the Taylor expansion of the accurate and surrogate forward models with respect to the uncertain parameter fields either as a control variate for variance reduction or as a means to directly and efficiently approximate the mean and covariance of the approximation errors. We propose efficient methods for evaluating the expressions for the mean and covariance of the Taylor approximations based on linear(-ized) PDE solves. Furthermore, the proposed approach is independent of the dimension of the uncertain parameter, depending instead on the intrinsic dimension of the data, ensuring scalability to high-dimensional problems. The potential benefits of the proposed approach are demonstrated for two high-dimensional inverse problems governed by PDE examples, namely for the estimation of a distributed Robin boundary coefficient in a linear diffusion problem, and for a coefficient estimation problem governed by a nonlinear diffusion problem.