Search NASA⌕ Search

SEARCH · Search NASA

Results for “reduced set methods”

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 505 records · Page 28

Multi-Level Structural Damage Characterization Using Sparse Acoustic Sensor Networks and Knowledge Transferred Deep Learning

Standard structural health monitoring techniques face well-known difficulties for comprehensive defect diagnosis in real-world structures that have structural, material, or geometric complexity. This motivates the exploration of machine-learning-based structural health monitoring methods in complex structures. However, creating sufficient training data sets with various defects is an ongoing challenge for data-driven machine (deep) learning algorithms. The ability to transfer the knowledge of a trained neural network from one component to another or to other sections of the same component would drastically reduce the required training data set. Also, it would facilitate computationally inexpensive machine learning based inspection systems. In this work, a machine-learning-based multi-level damage characterization is demonstrated with the ability to transfer trained knowledge within the sparse sensor network. A novel network spatial assistance and an adaptive convolution technique are proposed for efficient knowledge transfer within the deep learning algorithm. Proposed structural health monitoring method is experimentally evaluated on an aluminum plate with artificially induced defects. It was observed that the method improves the performance of knowledge transferred damage characterization by 50% during localization and 24% during severity assessment. Further, experiments using time windows with and without multiple edge reflections are studied. Results reveal that multiply scattered waves contain rich and deterministic defect signatures that can be mined using deep learning neural networks, improving the accuracy of both identification and quantification. In the case of a fixed sensor network, using multiply scattered waves shows 100% prediction accuracy at all levels of damage characterization.

36 MATERIALS SCIENCE↗

Learning nonlinear level sets for dimensionality reduction in function approximation

We developed a Nonlinear Level-set Learning (NLL) method for dimensionality reduction in high-dimensional function approximation with small data. This work is motivated by a variety of design tasks in real-world engineering applications, where practitioners would replace their computationally intensive physical models (e.g., high-resolution fluid simulators) with fast-to-evaluate predictive machine learning models, so as to accelerate the engineering design processes. There are two major challenges in constructing such predictive models: (a) high-dimensional inputs (e.g., many independent design parameters) and (b) small training data, generated by running extremely time-consuming simulations. Thus, reducing the input dimension is critical to alleviate the over-fitting issue caused by data insufficiency. Existing methods, including sliced inverse regression and active subspace approaches, reduce the input dimension by learning a linear coordinate transformation; our main contribution is to extend the transformation approach to a nonlinear regime. Specifically, we exploit reversible networks (RevNets) to learn nonlinear level sets of a high-dimensional function and parameterize its level sets in low-dimensional spaces. A new loss function was designed to utilize samples of the target functions' gradient to encourage the transformed function to be sensitive to only a few transformed coordinates. The NLL approach is demonstrated by applying it to three 2D functions and two 20D functions for showing the improved approximation accuracy with the use of nonlinear transformation, as well as to an 8D composite material design problem for optimizing the buckling-resistance performance of composite shells of rocket inter-stages.

Zhang, Guannan↗

Evolving bipartite authentication graph partitions

As large scale enterprise computer networks become more ubiquitous, finding the appropriate balance between user convenience and user access control is an increasingly challenging proposition. Suboptimal partitioning of users’ access and available services contributes to the vulnerability of enterprise networks. Previous edge-cut partitioning methods unduly restrict users’ access to network resources. This paper introduces a novel method of network partitioning superior to the current state-of-the-art which minimizes user impact by providing alternate avenues for access that reduce vulnerability. Networks are modeled as bipartite authentication access graphs and a multi-objective evolutionary algorithm is used to simultaneously minimize the size of large connected components while minimizing overall restrictions on network users. Lastly, results are presented on a real world data set that demonstrate the effectiveness of the introduced method compared to previous naive methods.

97 MATHEMATICS AND COMPUTING↗

Design and scale-up of 3D printed bat houses with biomass-derived polymer composites

Biomass (e.g., pine sawdust, especially high–ash content pine sawdust) is commonly disposed of as waste. Combining biomass with polymers to make composite feedstocks for 3D printing has been explored as a method to reduce or repurpose the biomass waste. Although not all biocomposite properties are known, the wood-based polylactic acid (PLA) composite has promising qualities for applications in ecological settings. In this work, pine wood–PLA composite feedstock was used to 3D print supplemental roost structures for endangered tree-roosting bats, which often face a paucity of suitable naturally occurring roosts. This material combination was selected because it is estimated to degrade faster than the synthetic material systems that are used widely in supplemental bat roosting structures to aid in the conservation of tree roosting bats. The layered, rough surface created by the 3D printing process serves as a surface that bats can grip while roosting. Computer-aided design (CAD) models were generated based on natural roost structures, and a full-size bat house was successfully additively manufactured using a pellet-fed large-scale 3D printing system. The 3D printed hexagon exhibited a tensile strength of 22–23 MPa and a Young’s modulus of 3202–3218 MPa in the x-direction. It has been demonstrated that the 3D printed bat house can be installed on a tree in a stable fashion. This successful demonstration of a bat roost manufactured using a bioderived composite should promote its use in other fish and wildlife structures and broader industrial applications such as construction and automobiles.

3D printing↗

Coherency-Constrained Spectral Clustering for Power Network Reduction

This paper presents a methodology for reducing the complexity of large-scale power network models using spectral clustering, aggregation of electrical components, and cost function approximation. Two approaches are explored using unconstrained and constrained spectral clustering to determine areas for effective system reduction. Once the system areas are determined, both loads and generators by type are aggregated, and their new cost function is approximated through polynomial curve-fitting or statistical methods. The performance of reduced networks is evaluated in terms of their ability to follow the true daily cost of the original system over a 24-hour period considering a set of several days. Two test systems are taken as test beds. Application of the methodology to a modified version of the IEEE 39-bus system reduces it from 17 generators to a 4-bus system and 9 generators with about 93% of accuracy. Similarly, the IEEE 118-bus system is reduced from 19 generators to a 3-bus system with three aggregated units achieving over 99% of accuracy. These findings address scalability challenges and enhance accuracy for high and mid-loading level conditions, and by aggregating thermal units with similar cost functions.

42 ENGINEERING↗

Multibeam field emission x-ray system with half-scan reconstruction algorithm

Purpose: In this article, the authors propose a multibeam field emission x-ray (MBFEX) system along with a half-scan fan-beam reconstruction algorithm. Methods: The proposed system consists of a linear CNT-based MBFEX source array, a single large area detector that is divided into same number of segments as the number of x-ray beams, a multihole collimator that aligns each beam with a corresponding detector segment, and a sample rotation stage. The collimator is placed between the source and the object to restrict the x-ray radiations through the target object only. In this design, all the x-ray beams are activated simultaneously to provide multiple projection views of the object. The detector is virtually segmented and synchronized with the x-ray exposure and the physiological signals when gating is involved. The transmitted x-ray intensity from each beam is collected by the corresponding segment on the detector. After each exposure, the object is rotated by a step angle until sufficient data set is collected. The half-scan reconstruction formula for MBFEX system is derived from the conventional filtered backprojection algorithm. To demonstrate the advantages of the system and method in reducing motion artifacts, the authors performed simulations with both standard and dynamic Shepp-Logan phantoms. Results: The numerical results indicate that the proposed multibeam system and the associated half-scan algorithm can effectively reduce the scanning time and improve the image quality for a time-varying object. Conclusions: The MBFEX technique offers an opportunity for the innovation of multisource imaging system.

60 APPLIED LIFE SCIENCES↗

Preliminary noise tests of the engine-over-the-wing concept. 2: 10 deg - 20 deg flap position

Preliminary acoustic tests of the engine-over-the-wing concept as a method for reducing the aerodynamic noise created by conventional and short takeoff aircraft are discussed. Tests were conducted with a small wing section model having two flaps which can be set for either the landing or takeoff positions. Data was acquired with the flaps set at 10 degrees and 20 degrees for takeoff and 30 and 60 degrees for landing. The engine exhaust was simulated by an air jet from a convergent nozzle. Far field noise data are presented for nominal pressure ratios of 1.25, 1.4 and 1.7 for both the flyover and sideline modes.

Reshotko, M.↗

On the evaluation of expected performance cost for partially observed closed-loop stochastic systems

New methods are presented for evaluating the expected performance cost of partially observed closed-loop stochastic systems. When the variances of the process statistics are small, a linearized model of the closed-loop stochastic system is defined for which the expected cost can be evaluated by recursion on a set of purely deterministic difference equations. When the variances of the process statistics are large, the linearized model can be used in the control variate method of variance reduction for reducing the number of sample paths required for effective Monte Carlo estimation.

Bayard, D. S.↗

The structure and energetics of Cr(CO)6 and Cr(CO)5

The geometric structure of Cr(CO)6 is optimized at the modified coupled pair functional (MCPF), single and double excitation coupled-cluster (CCSD) and CCSD(T) levels of theory (including a perturbational estimate for connected triple excitations), and the force constants for the totally symmetric representation are determined. The geometry of Cr(CO)5 is partially optimized at the MCPF, CCSD, and CCSD(T) levels of theory. Comparison with experimental data shows that the CCSD(T) method gives the best results for the structures and force constants, and that remaining errors are probably due to deficiencies in the one-particle basis sets used for CO. The total binding energies of Cr(CO)6 and Cr(CO)5 are also determined at the MCPF, CCSD, and CCSD(T) levels of theory. The CCSD(T) method gives a much larger total binding energy than either the MCPF or CCSD methods. An analysis of the basis set superposition error (BSSE) at the MCPF level of treatment points out limitations in the one-particle basis used. Calculations using larger basis sets reduce the BSSE, but the total binding energy of Cr(CO)6 is still significantly smaller than the experimental value, although the first CO bond dissociation energy of Cr(CO)6 is well described. An investigation of 3s3p correlation reveals only a small effect. In the largest basis set, the total CO binding energy of Cr(CO)6 is estimated to be 140 kcal/mol at the CCSD(T) level of theory, or about 86 percent of the experimental value. The remaining discrepancy between the experimental and theoretical value is probably due to limitations in the one-particle basis, rather than limitations in the correlation treatment. In particular an additional d function and an f function on each C and O are needed to obtain quantitative results. This is underscored by the fact that even using a very large primitive set (1042 primitive functions contracted to 300 basis functions), the superposition error for the total binding energy of Cr(CO)6 is 22 kcal/mol at the MCPF level of treatment.

Barnes, Leslie A.↗

A tool for manual endmember selection and spectral unmixing

Sampling a continuous radiance spectrum in many narrow contiguous spectral bands results in a high covariance between the bands. Hence, the true dimensionality of imaging spectrometer data is not determined by the number of spectral bands, but by the number of spectrally unique signatures whose mixtures reproduce the spectral variance observed in an image. Methods to unmix high dimensional multispectral data use principal components analysis to reduce the dimensionality. The variance of the spectral data is modeled as a linear combination of a finite set of endmembers in the space of the eigen-vectors that account for most of the variance. The number and characteristics of these endmembers are determined not only by the number and characteristics of the spectrally unique materials on the surface but also by processes (e.g., illumination, atmospheric scattering and absorption) affecting the signal received by the sensor. Selection of endmember spectra has typically been from a library. However, since most libraries are incomplete and do not account for the processes mentioned above, we have devised a computer display that allows researchers to explore interactively the eigenvector space of a representative and mean-corrected subset of the image data in search of extreme spectra to designate as endmembers. This display, which is based on parallel coordinates, is unique in the area of multidimensional visualization in that it includes not only a passive view of higher dimensional data but also the capability to interact and move geometrical objects in higher dimensional spaces.

Bateson, C. Ann↗

Mixed Integer Programming and Heuristic Scheduling for Space Communication

Optimal planning and scheduling for a communication network was created where the nodes within the network are communicating at the highest possible rates while meeting the mission requirements and operational constraints. The planning and scheduling problem was formulated in the framework of Mixed Integer Programming (MIP) to introduce a special penalty function to convert the MIP problem into a continuous optimization problem, and to solve the constrained optimization problem using heuristic optimization. The communication network consists of space and ground assets with the link dynamics between any two assets varying with respect to time, distance, and telecom configurations. One asset could be communicating with another at very high data rates at one time, and at other times, communication is impossible, as the asset could be inaccessible from the network due to planetary occultation. Based on the network's geometric dynamics and link capabilities, the start time, end time, and link configuration of each view period are selected to maximize the communication efficiency within the network. Mathematical formulations for the constrained mixed integer optimization problem were derived, and efficient analytical and numerical techniques were developed to find the optimal solution. By setting up the problem using MIP, the search space for the optimization problem is reduced significantly, thereby speeding up the solution process. The ratio of the dimension of the traditional method over the proposed formulation is approximately an order N (single) to 2*N (arraying), where N is the number of receiving antennas of a node. By introducing a special penalty function, the MIP problem with non-differentiable cost function and nonlinear constraints can be converted into a continuous variable problem, whose solution is possible.

Lee, Charles H.↗

Transconvolution and the virtual positron emission tomograph-A new method for cross calibration in quantitative PET/CT imaging

Purpose: Positron emission tomography (PET)/computed tomography (CT) measurements on small lesions are impaired by the partial volume effect, which is intrinsically tied to the point spread function of the actual imaging system, including the reconstruction algorithms. The variability resulting from different point spread functions hinders the assessment of quantitative measurements in clinical routine and especially degrades comparability within multicenter trials. To improve quantitative comparability there is a need for methods to match different PET/CT systems through elimination of this systemic variability. Consequently, a new method was developed and tested that transforms the image of an object as produced by one tomograph to another image of the same object as it would have been seen by a different tomograph. The proposed new method, termed Transconvolution, compensates for differing imaging properties of different tomographs and particularly aims at quantitative comparability of PET/CT in the context of multicenter trials. Methods: To solve the problem of image normalization, the theory of Transconvolution was mathematically established together with new methods to handle point spread functions of different PET/CT systems. Knowing the point spread functions of two different imaging systems allows determining a Transconvolution function to convert one image into the other. This function is calculated by convolving one point spread function with the inverse of the other point spread function which, when adhering to certain boundary conditions such as the use of linear acquisition and image reconstruction methods, is a numerically accessible operation. For reliable measurement of such point spread functions characterizing different PET/CT systems, a dedicated solid-state phantom incorporating {sup 68}Ge/{sup 68}Ga filled spheres was developed. To iteratively determine and represent such point spread functions, exponential density functions in combination with a Gaussian distribution were introduced. Furthermore, simulation of a virtual PET system provided a standard imaging system with clearly defined properties to which the real PET systems were to be matched. A Hann window served as the modulation transfer function for the virtual PET. The Hann's apodization properties suppressed high spatial frequencies above a certain critical frequency, thereby fulfilling the above-mentioned boundary conditions. The determined point spread functions were subsequently used by the novel Transconvolution algorithm to match different PET/CT systems onto the virtual PET system. Finally, the theoretically elaborated Transconvolution method was validated transforming phantom images acquired on two different PET systems to nearly identical data sets, as they would be imaged by the virtual PET system. Results: The proposed Transconvolution method matched different PET/CT-systems for an improved and reproducible determination of a normalized activity concentration. The highest difference in measured activity concentration between the two different PET systems of 18.2% was found in spheres of 2 ml volume. Transconvolution reduced this difference down to 1.6%. In addition to reestablishing comparability the new method with its parameterization of point spread functions allowed a full characterization of imaging properties of the examined tomographs. Conclusions: By matching different tomographs to a virtual standardized imaging system, Transconvolution opens a new comprehensive method for cross calibration in quantitative PET imaging. The use of a virtual PET system restores comparability between data sets from different PET systems by exerting a common, reproducible, and defined partial volume effect.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Stochastic density functional theory

Linear-scaling implementations of density functional theory (DFT) reach their intended efficiency regime only when applied to systems having a physical size larger than the range of their Kohn–Sham density matrix (DM). This causes a problem since many types of large systems of interest have a rather broad DM range and are therefore not amenable to analysis using DFT methods. For this reason, the recently proposed stochastic DFT (sDFT), avoiding exhaustive DM evaluations, is emerging as an attractive alternative linear-scaling approach. This review develops a general formulation of sDFT in terms of a (non)orthogonal basis representation and offers an analysis of the statistical errors (SEs) involved in the calculation. Using a new Gaussian-type basis-set implementation of sDFT, applied to water clusters and silicon nanocrystals, it demonstrates and explains how the standard deviation and the bias depend on the sampling rate and the system size in various types of calculations. We also develop a basis-set embedded-fragments theory, demonstrating its utility for reducing the SEs for energy, density of states and nuclear force calculations. Finally, we discuss the algorithmic complexity of sDFT, showing it has CPU wall-time linear-scaling. Lastly, the method parallelizes well over distributed processors with good scalability and therefore may find use in the upcoming exascale computing architectures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Transonic flow calculations

The development of relaxation methods to calculate transonic flows is discussed. Rather accurate predictions can be made for a number of flows of interest using the transonic potential flow equation, which may be derived from the Euler equations for inviscid compressible flow by introducing the assumption that the flow is irrotational. Once the choice of a mathematical model has been settled, the numerical procedure for actually computing a solution contains two main elements: the construction of a discrete approximation which converges to the solution of the continuous problem in the limit as the mesh width is reduced to zero, and the solution of the resulting set of nonlinear difference equations by a convergent iterative scheme. The choice of an appropriate coordinate system and its influence on the accuracy of the discrete approximation is discussed. Several applications of the general method are described.

Jameson, A.↗

Reducing the number of variational equations in the implementation of multiple shooting

The standard method of multiple shooting for a system of n first-order differential equations with k unknown initial conditions requires the integration of k sets of variational equations on the first shot and n sets of variational equations on every shot thereafter. This paper describes a variant of multiple shooting that requires the solution of k sets of variational equations on every shot. The technique applies to both linear and nonlinear boundary-value problems. Techniques to deal with difficulties unique to the solution of nonlinear problems are suggested.

Krogh, F. T.↗

A Comprehensive Comparison of Methods for Evaluating Dispatch of Long-Duration Energy Storage in Power Systems Models

Long-duration energy storage (LDES) could play a pivotal role in the transformation of electricity grids with high shares of variable renewable energy (VRE) such as solar and wind. However, the weather-dependent nature of VRE introduces challenges for grid balancing and stability, which LDES - along with short-duration energy storage (SDES) - can help address. However, modeling LDES in production cost models (PCMs) is particularly challenging due to the need for high temporal resolution over extended optimization windows while preserving chronology, which ensures the alignment of energy storage operations with VRE generation over multi-day periods. This report compares traditional dispatch methods with advanced LDES dispatch strategies, such as the extended horizon approach, across different PCM platforms and examines tradeoffs and scalability. The comparison reveals that the traditional 1-day optimization horizon within the PCM leads to inefficient utilization of LDES. In contrast, extending the optimization horizon as much as possible significantly reduces curtailment and improves storage dispatch, especially in renewable-dense systems. There is also promise in using state-of-charge or end volume targets set by an external model, however this requires an additional modeling set and generally increases computational burden. This paper presents a comparison of these various methods in a number of power systems, showing algorithms initially in small test systems and scaling up to large, country-wide simulations. Overall, the research presents the trade-offs of various computational methods and illustrates how LDES may play an essential role in power systems of the future.

14 SOLAR ENERGY↗

SU-F-J-39: Dose Reduction Strategy Using Attenuation-Based Tube Current Modulation Method in CBCT for IGRT

Purpose: To reduce radiation dose to the patients, tube current modulation (TCM) method has been actively used in diagnostic CT systems. However, TCM method has not yet been applied to a kV-CBCT system on a LINAC machine. The purpose of this study is to investigate whether the use of TCM method is desirable in kV-CBCT system for IGRT. We have developed an attenuation-based tube current modulation (a-TCM) method using the prior knowledge of treatment CT image of a patient. Methods: Patients go through a diagnostic CT scan for RT planning; therefore, using this prior information of CT images, one can estimate the total attenuation of an x-ray through the patient body in a CBCT setting for radiation therapy. We performed a numerical study incorporating major factors into account such as polychromatic x-ray, scatter, noise, and bow-tie filter to demonstrate that a-TCM method can produce equivalent quality of images at reduced imaging radiation doses. Using the CT projector program, 680 projection images of the pediatric XCAT phantom were obtained both in conventional scanning condition, i.e., without modulating the tube current, and in the proposed a-TCM scanning condition. FDK reconstruction algorithm was used for image reconstruction, and the organ dose due to imaging radiation has been calculated in both cases and compared using GATE/Geant4 simulation toolkit. Results: Reconstructed CT images in the a-TCM method showed similar SSIM values and noise properties to the reference images acquired by the conventional CBCT. In addition, reduction of organ doses ranged from 12% to 27%. Conclusion: We have successfully demonstrated the feasibility and dosimetric merit of the a-TCM method for kV-CBCT, and envision that it can be a useful option of CBCT scanning that provides patient dose reduction without degrading image quality.

60 APPLIED LIFE SCIENCES↗