Search NASA⌕ Search

SEARCH · Search NASA

Results for “hierarchical algorithm”

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 55 records · Page 3

Deep Hierarchical Super Resolution for Scientific Data

We present a novel technique for hierarchical super resolution (SR) with neural networks (NNs), which upscales volumetric data represented with an octree data structure to a high-resolution uniform gridwith minimal seam artifacts on octree node boundaries. Our method uses existing state-of-the-art SR models and adds flexibility to upscale input data with varying levels of detail across the domain, instead of only uniform grid data that are supported in previous approaches.The key is to use a hierarchy of SR NNs, each trained to perform 2x SR between two levels of detail, with a hierarchical SR algorithm that minimizes seam artifacts by starting from the coarsest level of detail and working up.We show that our hierarchical approach outperforms baseline interpolation and hierarchical upscaling methods, and demonstrate the usefulness of our proposed approach across three use cases including data reduction using hierarchical downsampling+SR instead of uniform downsampling+SR, computation savings for hierarchical finite-time Lyapunov exponent field calculation, and super-resolving low-resolution simulation results for a high-resolution approximation visualization.

97 MATHEMATICS AND COMPUTING↗

APHiD: Hierarchical Task Placement to Enable a Tapered Fat Tree Topology for Lower Power and Cost in HPC Networks

The power and procurement cost of bandwidth in system-wide networks has forced a steady drop in the byte/flop ratio. This trend of computation becoming faster relative to the network is expected to hold. In this paper, we explore how cost-oriented task placement enables reducing the cost of system-wide networks by enabling high performance even on tapered topologies where more bandwidth is provisioned at lower levels. We describe APHiD, an efficient hierarchical placement algorithm that uses new techniques to improve the quality of heuristic solutions and reduces the demand on high-level, expensive bandwidth in hierarchical topologies. We apply APHiD to a tapered fat-tree, demonstrating that APHiD maintains application scalability even for severely tapered network configurations. Using simulation, we show that for tapered networks APHiD improves performance by more than 50% over random placement and even 15% in some cases over costlier, state-of-the-art placement algorithms.

Michelogiannakis, George↗

Automatic optical detection and classification of marine animals around MHK converters using machine vision

Optical systems provide valuable information for evaluating interactions and associations between organisms and MHK energy converters and for capturing potentially rare encounters between marine organisms and MHK device. The deluge of optical data from cabled monitoring packages makes expert review time-consuming and expensive. We propose algorithms and a processing framework to automatically extract events of interest from underwater video. The open-source software framework consists of background subtraction, filtering, feature extraction and hierarchical classification algorithms. This principle classification pipeline was validated on real-world data collected with an experimental underwater monitoring package. An event detection rate of 100% was achieved using robust principal components analysis (RPCA), Fourier feature extraction and a support vector machine (SVM) binary classifier. The detected events were then further classified into more complex classes – algae | invertebrate | vertebrate, one species | multiple species of fish, and interest rank. Greater than 80% accuracy was achieved using a combination of machine learning techniques.

16 TIDAL AND WAVE POWER↗

Basker: Parallel sparse LU factorization utilizing hierarchical parallelism and data layouts

Transient simulation in circuit simulation tools, such as SPICE and Xyce, depend on scalable and robust sparse LU factorizations for efficient numerical simulation of circuits and power grids. As the need for simulations of very large circuits grow, the prevalence of multicore architectures enable us to use shared memory parallel algorithms for such simulations. A parallel factorization is a critical component of such shared memory parallel simulations. We develop a parallel sparse factorization algorithm that can solve problems from circuit simulations efficiently, and map well to architectural features. This new factorization algorithm exposes hierarchical parallelism to accommodate irregular structure that arise in our target problems. It also uses a hierarchical two-dimensional data layout which reduces synchronization costs and maps to memory hierarchy found in multicore processors. We present an OpenMP based implementation of the parallel algorithm in a new multithreaded solver called Basker in the Trilinos framework. Here, we present performance evaluations of Basker on the Intel SandyBridge and Xeon Phi platforms using circuit and power grid matrices taken from the University of Florida sparse matrix collection and from Xyce circuit simulation. Basker achieves a geometric mean speedup of 5.91× on CPU (16 cores) and 7.4× on Xeon Phi (32 cores) relative to state-of-the-art solver KLU. Basker outperforms Intel MKL Pardiso solver (PMKL) by as much as 30× on CPU (16 cores) and 7.5× on Xeon Phi (32 cores) for low fill-in circuit matrices. Furthermore, Basker provides 5.4× speedup on a challenging matrix sequence taken from an actual Xyce simulation.

97 MATHEMATICS AND COMPUTING↗

Anomaly Detection for Power System Generation Control based on Hierarchical DBSCAN

The generation level of a generator in power grid is under control of multiple control loops such as governor action, Automatic Generation Control (AGC), and manual control. These control loops are vulnerable to cyber attacks and could become potential targets for the malicious adversary. In this paper, we investigated the detection of abnormal generation controls induced by cyber attacks. One data-driven anomaly detection methodology is presented based on the behavior conformity of generation units located in the same balancing authority. A semi-supervised clustering algorithm with Hierarchical Density based Spatial Clustering of Application with Noise (HDBSCAN) is proposed for the detection model training. In our previous work, we obtained a synthetic dataset that covers scenarios including normal operation, generation control under ramp attack, switching attack, AGC integrity attacks, etc. and it is used to evaluate the proposed clustering algorithm. Experimental results show that the proposed algorithm provides better detection accuracy than K-means clustering and can distinguish not only between normal and abnormal generation controls but also among various anomaly scenarios.

24 POWER TRANSMISSION AND DISTRIBUTION↗

NOA: A Scalable Multi-Parent Clustering Hierarchy for WSNs

NOA is a multi-parent, N-tiered, hierarchical clustering algorithm that provides a scalable, robust and reliable solution to autonomous configuration of large-scale wireless sensor networks. The novel clustering hierarchy's inherent benefits can be utilized by in-network data processing techniques to provide equally robust, reliable and scalable in-network data processing solutions capable of reducing the amount of data sent to sinks. Utilizing a multi-parent framework, NOA reduces the cost of network setup when compared to hierarchical beaconing solutions by removing the expense of r-hop broadcasting (r is the radius of the cluster) needed to build the network and instead passes network topology information among shared children. NOA2, a two-parent clustering hierarchy solution, and NOA3, the three-parent variant, saw up to an 83% and 72% reduction in overhead, respectively, when compared to performing one round of a one-parent hierarchical beaconing, as well as 92% and 88% less overhead when compared to one round of two- and three-parent hierarchical beaconing hierarchy.

99 GENERAL AND MISCELLANEOUS↗

High performance FPGA embedded system for machine learning based tracking and trigger in sPhenix and EIC

We present a comprehensive end-to-end pipeline to classify triggers versus background events in this paper. This pipeline makes online decisions to select signal data and enables the intelligent trigger system for efficient data collection in the Data Acquisition System (DAQ) of the upcoming sPHENIX and future EIC (Electron-Ion Collider) experiments. Starting from the coordinates of pixel hits that are lightened by passing particles in the detector, the pipeline applies three-stage of event processing (hits clustering, track reconstruction, and trigger detection) and labels all processed events with the binary tag of trigger versus background events. The pipeline consists of deterministic algorithms such as clustering pixels to reduce event size, tracking reconstruction to predict candidate edges, and advanced graph neural network-based models for recognizing the entire jet pattern. In particular, we apply the message-passing graph neural network to predict links between hits and reconstruct tracks and a hierarchical pooling algorithm (DiffPool) to make the graph-level trigger detection. We obtain an impressive performance (≥70% accuracy) for trigger detection with only 3200 neuron weights in the end-to-end pipeline. We deploy the end-to-end pipeline into a field-programmable gate array (FPGA) and accelerate the three stages with speedup factors of 1152, 280, and 21, respectively.

Instruments & Instrumentation↗

An adaptive granular representative volume element model with an evolutionary periodic boundary for hierarchical multiscale analysis

The hierarchical multiscale analysis normally utilizes a microscopic representative volume element (RVE) model to capture path/history-dependent macroscopic responses instead of using phenomenological constitutive models. However, for problems involving large deformation, the current RVE model used in geomechanics may lose representative properties due to the progressive distortion of the RVE box, unless a proper reinitialization is applied. This work develops an adaptive RVE model in conjunction with an evolutionary periodic boundary (EPB) algorithm for hierarchical multiscale analysis of granular materials undergoing large deformation based on a recent RVE model proposed for coupling molecular dynamics and the material point method. The proposed adaptive RVE model avoids the reinitialization of the RVE box that even undergoes extremely large shear deformation; meanwhile, it accounts for the deformation history of the RVE model and treats the interaction between boundary particles and other image particles in a more efficient way. Numerical examples with extremely large deformation are used to illustrate the adaptive granular RVE model enhanced by the proposed EPB algorithm. Furthermore, some key features of this new methodology are further discussed for clarification.

42 ENGINEERING↗

Integrate Latimer Controls' Solution into RTAC (CRADA Final Report, CRD-23-24672)

Latimer Controls, Inc. was awarded two vouchers under the Department of Energy's American-Made Solar Prize Round 6 to conduct collaborative research at a national laboratory. The National Renewable Energy Laboratory (NREL) was selected as a partner to assist Latimer Controls in the performance evaluation of its photovoltaic (PV) control software. This collaboration focuses on developing a hardware-in-the-loop (HIL) testbed at NREL, which will be used to test and validate the Latimer PV control technology in a realistic yet de-risked environment. Both Latimer and NREL teams will work together to analyze the collected test data, derive insights, and disseminate the scientific findings. Recent studies underscore the potential of solar energy as a zero-marginal-cost and zero-emission flexibility resource within the bulk power system, particularly when integrated with advanced control systems. To enhance the performance of such systems, Latimer Controls has developed leading-edge technologies, including machine learning (ML) algorithms and hierarchical inverter set-point allocation methods. These innovations are designed to estimate the operational headroom of large PV plants for grid integration and control. However, comprehensive validation under real-world conditions remains necessary. To address this gap, the concurrent CRADA project proposes the real-world application and validation of the Latimer Control solution within a HIL environment. Initially, the Latimer algorithm was developed and tested within MATLAB Simulink, a platform suitable for research-level simulations and iterative development. However, transitioning this technology to a real solar site as an industry-ready solution necessitates implementation in a format compatible with widely used solar power plant controllers. In this additional CRADA work, the MATLAB Simulink-based logic will be translated into Structured Text, a programming language compliant with IEC 61131 standards, which is commonly used for custom logic implementations in industry-leading programmable logic controllers (PLCs), such as the Schweitzer SEL real-time automation controller (RTAC). This transition will facilitate the deployment of the Latimer Control solution in real-world solar power plants, thereby advancing the technology towards commercialization.

14 SOLAR ENERGY↗

Optimizing group structures using hierarchical division

Creating group structures with few groups that give low errors is a difficult problem in reactor analysis. In recent years, automated optimization techniques have been applied to this task. We continue this trend by applying the hierarchical division algorithm to generate optimized group structures that minimize a cost function. At each stage, the algorithm adds a single group boundary to an existing group structure, dividing one group into two to increase the resolution of the group structure. The location of the added boundary is the one that gives the lowest error over all possible new boundary locations. Our implementation requires a beginning group structure, a set of candidate new boundary locations, and a set of reference reaction rates. As a proof of concept, we used WIMS-69 as the initial group structure, XMAS-172 as the ending group structures, and a 344-group reference group structure. Testing on two simple, homogenized reactor problems, we found that hierarchical division was able to reduce the error by a factor of around 5 with an increase of only 15% in the number of groups. Because hierarchical division can get stuck in local minima, it often reaches a plateau in its error reduction capability as many groups are added. Nevertheless, we find hierarchical division has strong potential to make good group structures into great group structures at a modest increase in computational cost. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Rapid Construction of Fe–Co–Ni Composition-Phase Map by Combinatorial Materials Chip Approach

100 nm thick Fe-Co-Ni materials chips were prepared and isothermally annealed at 500, 600 and 700oC, respectively. Pixel-by-pixel composition and structural mapping was performed by micro-beam X-ray at synchrotron light source. Diffraction images were recorded at a rate of 1 pattern/s. The XRD patterns were automatically processed, phase-identified and categorized by hierarchical clustering algorithm to construct the composition-phase map. Furthermore, the resulting maps are consistent with corresponding isothermal sections reported in the ASM Alloy Phase Diagram DatabaseTM, verifying the effectiveness of the present approach in phase diagram construction.

36 MATERIALS SCIENCE↗

Phlex: Parallel, Hierarchical, and Layered EXecution of data-processing algorithms

Phlex is a computing framework supporting the parallel, hierarchical, and layered execution of data-processing algorithms. It is based on the functional-programming paradigm, thus guaranteeing thread-safety when invoking user-defined pure functions. Phlex allows users to specify arbitrary graph-based hierarchies of data organization, enabling more flexible processing of data as required by the constraints of the program.

Knoepfel, KyleJ. [Fermi National Accelerator Labor↗

An efficient algorithm for generating diverse microstructure sets and delineating properties closures

Properties closures delineate the theoretical objective space for materials design problems, allowing designers to make informed trade-offs between competing constraints and target properties. In this paper, we present a new algorithm called hierarchical simplex sampling (HSS) that approximates properties closures more efficiently and faithfully than traditional optimization based approaches. By construction, HSS generates samples of microstructure statistics that span the corresponding microstructure hull. As a result, we also find that HSS can be coupled with synthetic polycrystal generation software to generate diverse sets of microstructures for subsequent mesoscale simulations. Finally, by more broadly sampling the space of possible microstructures, it is anticipated that such diverse microstructure sets will expand our understanding of the influence of microstructure on macroscale effective properties and inform the construction of higher-fidelity mesoscale structure-property models.

36 MATERIALS SCIENCE↗

A graph-based approach to developing adaptive representations of complex reaction mechanisms

An effective adaptive mechanism reduction approach based on flux graph clustering is proposed in this paper. The instantaneous element flux is quantified and considered as a proxy for describing the reactive propensities of the system. Our underlying hypothesis is that even though particular conditions may be characterized by a multitude of combinations of species mass fraction, T, and P, the essential chemistry, and hence the reaction propensity of the mixture that is active under this family of conditions, is the same. Therefore, we opt to use the instantaneous fluxes through the active reactions as an intrinsic property of the system. Flux graphs are first constructed for the chemical reaction system under numerous conditions aiming at capturing the attainable region. Similarity between flux graphs is quantified through the distances between corresponding vectors, using the cosine coefficient and a novel graph-distance metric taking into account the magnitude of each flux and the activity distribution of different fluxes. A hierarchical clustering algorithm is implemented to group similar instantaneous flux graphs into clusters, and consequently a reduced mechanism is generated for each cluster. A search algorithm is defined afterward to assign a new query point to a particular flux graph cluster, and subsequently the reduced mechanism associated with this cluster is used to describe the system at this time point. Finally, the methodology is demonstrated using n-pentane combustion in an adiabatic plug flow reactor model and a pairwise mixing stirred reactor model. (author)

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Enabling machine learning-ready HPC ensembles with Merlin

With the growing complexity of computational and experimental facilities, many scientific researchers are turning to machine learning (ML) techniques to analyze large scale ensemble data. With complexities such as multi-component workflows, heterogeneous machine architectures, parallel file systems, and batch scheduling, care must be taken to facilitate this analysis in a high performance computing (HPC) environment. Here, we present Merlin, a workflow framework to enable large ML-friendly ensembles of scientific HPC simulations. By augmenting traditional HPC with distributed compute technologies, Merlin aims to lower the barrier for scientific subject matter experts to incorporate ML into their analysis. As a producer–consumer workflow model, Merlin enables multi-machine, cross-batch job, dynamically allocated yet persistent workflows capable of utilizing surge-compute resources. Key features of Merlin are a flexible HPC-centric interface, low per-task overhead, multi-tiered fault recovery, and a hierarchical sampling algorithm that allows for $\mathscr{O}$(N) task execution and $\mathscr{O}$(N ln N) task queuing to ensembles of millions of tasks. In addition to Merlin’s design, we test the algorithm’s performance in an HPC center and demonstrate the ability to enqueue 40 million simulations in 100 s, with a 30 millisecond per-task overhead that is independent of ensemble size. Finally, we describe some example applications that Merlin has enabled on leadership-class HPC resources, such as the ML-augmented optimization of nuclear fusion experiments and the calibration of infectious disease models to study the progression of and possible mitigation strategies for COVID-19.

97 MATHEMATICS AND COMPUTING↗

Choreographing Couch and Collimator in Volumetric Modulated Arc Therapy

Purpose: To design and optimize trajectory-based, noncoplanar subarcs for volumetric modulated arc therapy (VMAT) deliverable on both Varian TrueBEAM system and traditional accelerators; and to investigate their potential advantages for treating central nervous system (CNS) tumors. Methods and Materials: To guide the computerized selection of beam trajectories consisting of simultaneous couch, gantry, and collimator motion, a score function was implemented to estimate the geometric overlap between targets and organs at risk for each couch/gantry angle combination. An initial set of beam orientations is obtained as a function of couch and gantry angle, according to a minimum search of the score function excluding zones of collision. This set is grouped into multiple continuous and extended subarcs subject to mechanical limitations using a hierarchical clustering algorithm. After determination of couch/gantry trajectories, a principal component analysis finds the collimator angle at each beam orientation that minimizes residual target-organ at risk overlaps. An in-house VMAT optimization algorithm determines the optimal multileaf collimator position and monitor units for control points within each subarc. A retrospective study of 10 CNS patients compares the proposed method of VMAT trajectory with dynamic gantry, leaves, couch, and collimator motion (Tra-VMAT); a standard noncoplanar VMAT with no couch/collimator motion within subarcs (Std-VMAT); and noncoplanar intensity-modulated radiotherapy (IMRT) plans that were clinically used. Results: Tra-VMAT provided improved target dose conformality and lowered maximum dose to brainstem, optic nerves, and chiasm by 7.7%, 1.1%, 2.3%, and 1.7%, respectively, compared with Std-VMAT. Tra-VMAT provided higher planning target volume minimum dose and reduced maximum dose to chiasm, optic nerves, and cochlea by 6.2%, 1.3%, 6.3%, and 8.4%, respectively, and reduced cochlea mean dose by 8.7%, compared with IMRT. Tra-VMAT averaged beam-on time was comparable to Std-VMAT but significantly (45%) less than IMRT. Conclusion: Optimized couch, gantry, and collimator trajectories may be integrated into VMAT with improved mechanical flexibility and may provide better dosimetric properties and improved efficiency in the treatment of CNS tumors.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Coherent correlation imaging for resolving fluctuating states of matter

Fluctuations and stochastic transitions are ubiquitous in nanometre-scale systems, especially in the presence of disorder. However, their direct observation has so far been impeded by a seemingly fundamental, signal-limited compromise between spatial and temporal resolution. Here we develop coherent correlation imaging (CCI) to overcome this dilemma. Our method begins by classifying recorded camera frames in Fourier space. Contrast and spatial resolution emerge by averaging selectively over same-state frames. Temporal resolution down to the acquisition time of a single frame arises independently from an exceptionally low misclassification rate, which we achieve by combining a correlation-based similarity metric with a modified, iterative hierarchical clustering algorithm. We apply CCI to study previously inaccessible magnetic fluctuations in a highly degenerate magnetic stripe domain state with nanometre-scale resolution. We uncover an intricate network of transitions between more than 30 discrete states. Our spatiotemporal data enable us to reconstruct the pinning energy landscape and to thereby explain the dynamics observed on a microscopic level. CCI massively expands the potential of emerging high-coherence X-ray sources and paves the way for addressing large fundamental questions such as the contribution of pinning and topology in phase transitions and the role of spin and charge order fluctuations in high-temperature superconductivity.

36 MATERIALS SCIENCE↗

HYPERCOMPACT STELLAR SYSTEMS AROUND RECOILING SUPERMASSIVE BLACK HOLES

A supermassive black hole ejected from the center of a galaxy by gravitational-wave recoil carries a retinue of bound stars - a 'hypercompact stellar system' (HCSS). The numbers and properties of HCSSs contain information about the merger histories of galaxies, the late evolution of binary black holes, and the distribution of gravitational-wave kicks. We relate the structural properties (size, mass, density profile) of HCSSs to the properties of their host galaxies and to the size of the kick in two regimes: collisional (M {sub BH} {approx}< 10{sup 7} M {sub sun}), i.e., short nuclear relaxation times, and collisionless (M {sub BH} {approx}> 10{sup 7} M {sub sun}), i.e., long nuclear relaxation times. HCSSs are expected to be similar in size and luminosity to globular clusters, but in extreme cases (large galaxies, kicks just above escape velocity) their stellar mass can approach that of ultracompact dwarf galaxies. However, they differ from all other classes of compact stellar system in having very high internal velocities. We show that the kick velocity is encoded in the velocity dispersion of the bound stars. Given a large enough sample of HCSSs, the distribution of gravitational-wave kicks can therefore be empirically determined. We combine a hierarchical merger algorithm with stellar population models to compute the rate of production of HCSSs over time and the probability of observing HCSSs in the local universe as a function of their apparent magnitude, color, size, and velocity dispersion, under two different assumptions about the star formation history prior to the kick. We predict that {approx}10{sup 2} HCSSs should be detectable within 2 Mpc of the center of the Virgo cluster, and that many of these should be bright enough that their kick velocities (i.e., velocity dispersions) could be measured with reasonable exposure times. We discuss other strategies for detecting HCSSs and speculate on some exotic manifestations.

79 ASTRONOMY AND ASTROPHYSICS↗