Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “clustering 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
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Data-Driven Performance Optimization of Gamma Spectrometers With Many Channels

In gamma spectrometers with variable spectroscopic performance across many channels (e.g., many pixels or voxels), a tradeoff exists between including data from successively worse-performing readout channels and increasing efficiency. Brute-force calculation of the optimal set of included channels is exponentially infeasible as the number of channels grows, and approximate methods are required. In this work, we present a data-driven framework for attempting to find near-optimal sets of included detector channels. The framework leverages non-negative matrix factorization (NMF) to learn the behavior of gamma spectra across the detector and clusters similarly-performing detector channels together. Performance comparisons are then made between spectra with channel clusters removed, which is more feasible than brute force. The framework is general and can be applied to arbitrary, user-defined performance metrics depending on the application. We apply this framework to optimizing gamma spectra measured by H3D M400 CdZnTe (CZT) spectrometers, which exhibit variable performance across their crystal volumes. In particular, we show several examples optimizing various performance metrics for uranium and plutonium gamma spectra in non-destructive assay (NDA) for nuclear safeguards, and explore trends in performance versus parameters such as clustering algorithm type. We also compare the NMF + clustering pipeline to several non-machine-learning (ML) algorithms, including several greedy algorithms. Although, we find that the NMF + clustering pipeline tends to find the best-performing set of detector voxels, significantly improving over the unoptimized spectra, but that a greedy accumulation of spectra segmented by detector depth can, in some cases, give similar performance improvements in much less computation time.

Energy resolution↗
DOE OSTI2023

Leveraging the digital thread for physics-based prediction of microstructure heterogeneity in additively manufactured parts

A major limitation of additive manufacturing (AM) processes is that local conditions of material deposition frequently lead to unintentional heterogeneities in microstructure and properties within a single component, despite nominally uniform process conditions. Up to now, there has been no way to a priori determine the distribution of these heterogeneities, requiring expensive trial-and-error approaches to fabrication, testing, and characterization. Here, a physics-based framework for creating a digital representation of the laser powder bed fusion (PBF) process is proposed to predict the variation in solidification behavior that leads to heterogeneous microstructures in an as-built part. By leveraging in situ process data stored in the part’s digital thread, the scan path and process parameters were input into a heat transfer model which predicted solidification data at the melt pool scale. A two-step unsupervised clustering algorithm was used to first cluster the local solidification conditions (12.5µm 3 voxels) and then to cluster the regional behavior on the scale of multiple scan passes and print layers (250µm 3 super-voxels). This process was used to identify regions with similar solidification characteristics for multiple locations in a Stainless Steel 316-L component. The corresponding as-built part was sectioned and characterized using electron backscatter diffraction (EBSD). Quantitative analysis of the pole figures confirmed that the predicted regions of heterogeneity in the solidification conditions corresponded with differences in the observed microstructure. In conclusion, this work shows a viable path for estimating the microstructural heterogeneity for additively manufactured parts to either limit microstructural variation throughout a part or to enable functionality-based variation of the microstructure.

36 MATERIALS SCIENCE↗
DOE OSTI2025

ArborX 2.0

ArborX library tackles a problem of efficiently finding geometric objects that are close in space. Variations of this problem, such as finding the nearest neighbors of a point, or finding all objects within a certain distance, are inherent components of applications in many fields. The data may be large so that solving the problem efficiently may require significant computational resources, such as multiple processors or accelerators such as general purpose GPUs. ArborX' main advantage in its ability to solve large problems efficiently utilizing a combination of distributed and on-node parallelism. ArborX can be run efficiently on a wide variety of hardware, including GPUs from different vendors, which distinguishes it from other available libraries which typically choose only few of these. The other advantage is that it supports both types of user problems: spatial problems (useful for intersections and finding objects within certain distance), and nearest neighbor problems. ArborX also supports flexible interface in its interaction with a user. Particularly, it allows a user to call user's own function on a positive match, a functionality not rarely available in other libraries. ArborX implements construction and traversal algorithms using efficient tree structures, such as bounding volume hierarchy (BVH). At its core, ArborX uses linear BVH for its low construction cost and sufficient quality. ArborX implements both spatial and nearest-neighbor traversal algorithms. ArborX also provides several clustering algorithms (minimum spanning tree, DBSCAN, HDBSCAN*), interpolation using minimum least squares and ray tracing. ArborX is written using C++, and is parallelized using the message passing interface (MPI) for the distributed communication, and the Kokkos library for on-node parallelism. This approach allows ArborX to be run on a wide variety of hardware, from common laptops and desktops to supercomputers while using the same codebase.

Prokopenko, Andrey [Oak Ridge National Laboratory ↗
DOE OSTI2021

Automatic Traffic Queue-End Identification using Location-Based Waze User Reports

Traffic queues, especially queues caused by non-recurrent events such as incidents, are unexpected to high-speed drivers approaching the end of queue (EOQ) and become safety concerns. Though the topic has been extensively studied, the identification of EOQ has been limited by the spatial-temporal resolution of traditional data sources. This study explores the potential of location-based crowdsourced data, specifically Waze user reports. It presents a dynamic clustering algorithm that can group the location-based reports in real time and identify the spatial-temporal extent of congestion as well as the EOQ. The algorithm is a spatial-temporal extension of the density-based spatial clustering of applications with noise (DBSCAN) algorithm for real-time streaming data with an adaptive threshold selection procedure. Here, the proposed method was tested with 34 traffic congestion cases in the Knoxville, Tennessee area of the United States. It is demonstrated that the algorithm can effectively detect spatial-temporal extent of congestion based on Waze report clusters and identify EOQ in real-time. The Waze report-based detection are compared to the detection based on roadside sensor data. The results are promising: The EOQ identification time of Waze is similar to the EOQ detection time of traffic sensor data, with only 1.1 min difference on average. In addition, Waze generates 1.9 EOQ detection points every mile, compared to 1.8 detection points generated by traffic sensor data, suggesting the two data sources are comparable in respect of reporting frequency. The results indicate that Waze is a valuable complementary source for EOQ detection where no traffic sensors are installed.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2023

Fast tree-based algorithms for DBSCAN for low-dimensional data on GPUs

DBSCAN is a well-known density-based clustering algorithm to discover arbitrary shape clusters. While conceptually simple in serial, the algorithm is challenging to efficiently parallelize on manycore GPU architectures. Common pitfalls, such as asynchronous range query calls, result in high thread execution divergence in many implementations. In this paper, we propose a new framework for GPU-accelerated DBSCAN, and describe two tree-based algorithms within that framework. Both algorithms fuse the search for neighbors with updating cluster information, but differ in their treatment of dense regions of the data. We show that the time taken to compute clusters is at most twice that of determination of the neighbors. We compare the proposed algorithms with existing CPU and GPU implementations, and demonstrate their competitiveness and performance using a fast traversal structure (bounding volume hierarchy) for low dimensional data. We also show that the memory usage can be reduced by processing object neighbors dynamically without storing them.

Prokopenko, Andrey↗
DOE OSTI2022

Machine-learning-based automatic small-angle measurement between planar surfaces in interferometer images: A 2D multilayer Laue lenses case

Here, we report a new machine-learning-based approach to automatically measure the small angle between multiple planar surfaces characterized by white light interferometers. By applying an unsupervised clustering algorithm, DBSCAN (Density-Based Spatial Clustering of Applications with Noise), the multiple surfaces in an interferometer image are automatically identified as distinct surfaces. The angles between every two surfaces are then calculated through the surface fitting. This method can be applied to multiple surfaces regardless of their shapes and locations and significantly simplifies the angle measurement procedure. Using the developed method, we have demonstrated a quick and precise angle measurement for the alignment of 2D Multilayer Laue Lenses (MLLs) for the development of high-resolution x-ray microscopy. This automatic, accurate, and robust small-angle measurement method is compatible with widely used white light interferometers and can be further applied to other metrology applications of interferometer results.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Harnessing the predicted maize pan-interactome for putative gene function prediction and prioritization of candidate genes for important traits

Abstract The recent assembly and annotation of the 26 maize nested association mapping population founder inbreds have enabled large-scale pan-genomic comparative studies. These studies have expanded our understanding of agronomically important traits by integrating pan-transcriptomic data with trait-specific gene candidates from previous association mapping results. In contrast to the availability of pan-transcriptomic data, obtaining reliable protein–protein interaction (PPI) data has remained a challenge due to its high cost and complexity. We generated predicted PPI networks for each of the 26 genomes using the established STRING database. The individual genome-interactomes were then integrated to generate core- and pan-interactomes. We deployed the PPI clustering algorithm ClusterONE to identify numerous PPI clusters that were functionally annotated using gene ontology (GO) functional enrichment, demonstrating a diverse range of enriched GO terms across different clusters. Additional cluster annotations were generated by integrating gene coexpression data and gene description annotations, providing additional useful information. We show that the functionally annotated PPI clusters establish a useful framework for protein function prediction and prioritization of candidate genes of interest. Our study not only provides a comprehensive resource of predicted PPI networks for 26 maize genomes but also offers annotated interactome clusters for predicting protein functions and prioritizing gene candidates. The source code for the Python implementation of the analysis workflow and a standalone web application for accessing the analysis results are available at https://github.com/eporetsky/PanPPI.

Genetics & Heredity↗
DOE OSTI2024

Identifying Climate Patterns Using Clustering Autoencoder Techniques

Abstract The complexity of growing spatiotemporal resolution of climate simulations produces a variety of climate patterns under different projection scenarios. This paper proposes a new data-driven climate classification workflow via an unsupervised deep learning technique that can dimensionally reduce the vast volume of spatiotemporal numerical climate projection data into a compact representation. We aim to identify distinct zones that capture multiple climate variables as well as their future changes under different climate change scenarios. Our approach leverages convolutional autoencoders combined with k -means clustering (standard autoencoder) and online clustering based on the Sinkhorn–Knopp algorithm (clustering autoencoder) across the conterminous United States (CONUS) to capture unique climate patterns in a data-driven fashion from the Geophysical Fluid Dynamics Laboratory Earth System Model with GOLD component (GFDL-ESM2G). The developed approach compresses 70 years of GFDL-ESM2G simulation at 0.125° spatial resolution across the CONUS under multiple warming scenarios to a lower-dimensional space by a factor of 660 000 and then tested on 150 years of GFDL-ESM2G simulation data. The results show that five climate clusters capture physically reasonable and spatially stable climatological patterns matched to known climate classes defined by human experts. Results also show that using a clustering autoencoder can reduce the computational time for clustering by up to 9.2 times when compared to using a standard autoencoder. Our five unique climate patterns resulting from the deep learning–based clustering of the lower-dimensional space thereby enable us to provide insights on hydrometeorology and its spatial heterogeneity across the conterminous United States immediately without downloading large climate datasets. Significance Statement This paper presents a data-driven climate classification approach using unsupervised deep learning to dimensionally reduce climate model outputs and to identify distinct climate regions for their future changes. Our approach compresses climate information for 70 years of Geophysical Fluid Dynamics Laboratory Earth System Model data across the conterminous United States (CONUS) at 0.125° spatial resolution. The results reveal that five climate clusters capture reasonable and stable climatological patterns matched to known climate patterns. The embedded clustering process in deep learning provides ×9.2 times faster execution than the k -means clustering technique. These results give us insight about climate spatial patterns and heterogeneity of hydrological patterns across the conterminous United States without downloading large climate datasets.

Kurihana, Takuya↗
DOE OSTI2024

PANDORA: A Parallel Dendrogram Construction Algorithm for Single Linkage Clustering on GPU

This paper introduces Pandora, a parallel algorithm for computing dendrograms, the hierarchical cluster trees for single linkage clustering (SLC). Current parallel approaches construct dendrograms by partitioning a minimum spanning tree and removing edges. However, they struggle with skewed, hard-to-parallelize real-world dendrograms. Consequently, computing dendrograms is the sequential bottleneck in HDBSCAN*[21], a popular SLC variant. Pandora uses recursive tree contraction to address this limitation. Pandora contracts nodes to construct progressively smaller trees. It computes the smallest contracted dendrogram and expands it by inserting contracted edges. This recursive strategy is highly parallel, skew-independent, work-optimal, and well-suited for GPUs and multicores. We develop a performance portable implementation of Pandora in Kokkos[31] and evaluate its performance on multicore CPUs and multi-vendor GPUs (e.g., Nvidia, AMD) for dendrogram construction in HDBSCAN*. Multithreaded Pandora is 2.2x faster than the current best-multithreaded implementation. Our GPU version achieves 6-20x speedup on AMD GPUs and 10-37x on NVIDIA GPUs over multithreaded Pandora. Pandora removes HDBSCAN*’s sequential bottleneck, greatly boosting efficiency, particularly with GPUs.

Sao, Piyush↗
DOE OSTI2025

Distinguishing isotropic and anisotropic signals for X-ray total scattering using machine learning

Understanding structure–property relationships is essential for advancing technologies based on thin films. X-ray pair distribution function (PDF) analysis can access relevant atomic structure details spanning local-, mid- and long-range structure. While X-ray PDF has been adapted for thin films on amorphous substrates, measurements on single-crystal substrates are necessary to accurately determine structure origins for some thin film materials, especially those for which the substrate changes the accessible structure and properties. However, when measuring films on single-crystal substrates, high-intensity anisotropic Bragg spots saturate 2D detector images, overshadowing the thin films' isotropic scattering signal. This renders previous data processing methods for films on amorphous substrates unsuitable for films on single-crystal substrates. To address this measurement need, we developed IsoDAT2D, an innovative data processing approach using unsupervised machine learning algorithms. The program combines dimensionality reduction and clustering algorithms to separate thin film and single-crystal substrate X-ray scattering signals. We use SimDAT2D , a program we developed to generate simulated thin film data, to validate IsoDAT2D . Here we also use IsoDAT2D to isolate X-ray total scattering signal from a thin film on a single-crystal substrate. The resulting PDF data are compared with similar data processed using previous methods, especially substrate subtraction for single-crystal and amorphous substrates. PDF data from IsoDAT2D -identified X-ray total scattering data are significantly better than from single-crystal substrate subtraction, but not as reliable as PDF data from amorphous substrate subtraction. With IsoDAT2D , there are new opportunities to expand PDF to a wider variety of thin films, including those on single-crystal substrates, with which new structure–property relationships can be elucidated to enable fundamental understanding and technological advances.

36 MATERIALS SCIENCE↗
DOE OSTI2024

A network approach for multiscale catchment classification using traits

Abstract. The classification of river catchments into groups with similar biophysical characteristics is useful to understand and predict their hydrological behavior. The increasing availability of remote sensing and other large-scale geospatial datasets has enabled the use of advanced data-driven approaches to classify catchments using traits such as topography, geology, climate, land cover, land use, and human influence. Unsupervised clustering algorithms based on the Euclidean distance are commonly used for trait-based classification but are not suitable for highly dimensional data. In this study we present a new network-based method for multi-scale catchment classification, which can be applied to large datasets and used to determine the traits associated with different catchment groups. In this framework, two networks are analyzed in parallel: the first being where the nodes are traits and the second being where the nodes are catchments. In both cases, edges represent pairwise similarity, and a network cluster detection algorithm is used for the classification. The trait network is used to investigate redundancy in the trait data and to condense this information into a small number of interpretable categories. The catchments network is used to classify the catchments into clusters and to identify representative catchments for the different groups using the degree centrality metric. We apply this method to classify 9067 river catchments across the contiguous United States at both regional and continental scales using 274 non-categorical traits. At the continental scale, we identify 25 interpretable trait categories and 34 catchment clusters of sizes greater than 50. We find that catchments with similar trait categories are typically located in the same region, with different spatial patterns emerging among clusters dominated by natural and anthropogenic traits. We also find that the catchment clusters exhibit distinct hydrological behavior based on an analysis of streamflow indices. This network approach provides several advantages over traditional means of classification, including better separation of clusters, the use of alternate similarity metrics that are more suitable for highly dimensional data, and reducing redundancy in the trait information. The paired catchment–trait networks enable analysis of hydrological behavior using the dominant trait categories for each catchment cluster. The approach can be used at multiple spatial scales since the network topologies adjust automatically to reflect the trait patterns at the scale of investigation. Finally, the representative catchments identified as hub nodes in the network can be used to guide transferable observational and modeling strategies. The method is broadly applicable beyond hydrology for classification of other complex systems that utilize different types of trait datasets.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

A Hybrid DC Fault Primary Protection Algorithm for Multi-Terminal HVdc Systems

Protection against dc faults is one of the main technical hurdles faced when operating converter-based HVdc systems. Protection becomes even more challenging for multi-terminal dc (MTdc) systems with more than two terminals/converter stations. In this paper, a hybrid primary fault detection algorithm for MTdc systems is proposed to detect a broad range of failures. Sensor measurements, i.e., line currents and dc reactor voltages measured at local terminals, are first processed by a top-level context clustering algorithm. For each cluster, the best fault detector is selected among a detector pool according to a rule resulting from a learning algorithm. The detector pool consists of several existing detection algorithms, each performing differently across fault scenarios. The proposed hybrid primary detection algorithm: i) offers superior performance compared to an individual detector through a data-driven approach; ii) detects all major fault types including pole-to-pole (P2P), pole-to-ground (P2G), and external dc faults; iii) identifies faults with various fault locations and impedances; iv) is more robust to noisy sensor measurements compared to existing methods; v) does not require exhaustive simulation and sampling for training the model. Performance and effectiveness of the proposed algorithm are evaluated and verified based on time-domain simulations in the PSCAD/EMTDC software environment. The results confirm satisfactory operation, accuracy, and detection speed of the proposed algorithm under various fault scenarios.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2020

Hypergraph Random Walks, Laplacians, and Clustering

We propose a flexible framework for clustering hypergraph-structured data based on recently proposed random walks utilizing edge-dependent vertex weights. When incorporating edge-dependent vertex weights (EDVW), a weight is associated with each vertex-hyperedge pair, yielding a weighted incidence matrix of the hypergraph. Such weightings have been utilized in term-document representations of text data sets. We explain how random walks with EDVW serve to construct different hypergraph Laplacian matrices, and then develop a suite of clustering methods that use these incidence matrices and Laplacians for hypergraph clustering. Using 20Newsgroup, U.S. patent, Reuters' Corpus Volume 1, and genetics data sets, we compare the performance of these clustering algorithms experimentally against a variety of existing hypergraph clustering methods. We show that the proposed methods produce higher-quality clusters.

hypergraphs, clustering, laplacian, random walk, M↗
DOE OSTI2024

TICC Clustering Library v.1.0

SAND2024-01234O TICC is a clustering algorithm that labels a sequence of data points according to numerical properties. This library is a Python implementation of the algorithm described in "Toeplitz Inverse Covariance-Based Clustering of Multivariate Time Series Data" (Hallac et al. 2017). It includes documentation, performance improvements, examples, and test coverage. This library allows users to automatically segment a series of multivariate data points according to their covariance—that is, the way the values at each data point are changing in relation to one another. This is useful for identifying periods in which a system is behaving. For example, if a sensor is measuring a car's velocity, steering wheel angle, braking and acceleration, TICC can determine when the car was stopped, beginning/exiting a turn, slowing or accelerating at an intersection, or driving on straight or curved roads. TICC can be applied to measure multiple quantities at known times. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525

Dalbey, Keith↗
DOE OSTI2021

Secondary Crash Identification using Crowdsourced Waze User Reports

Secondary crashes are crashes that occur as a result of the nonrecurrent congestion originating from primary crashes, and always have a greater impact on safety and traffic than a single crash. A better understanding of secondary crashes would benefit traffic incident management, and this requires accurate identification of secondary crashes. This study explores using crowdsourced Waze user reports to identify secondary crashes. Here, a network-based clustering algorithm is proposed to extract the primary crash cluster, including all user reports originating from the primary crash, and any crash that occurred within the cluster would be a secondary crash. This method works as a filter to select accurate primary–secondary relationships, thus precisely identifying secondary crashes. A case study is performed with crashes occurring from June to December 2019 on a 30-mi stretch of I-40 in Knoxville, TN. A static threshold method (crash duration and 10 mi) was used to preselect the potential primary–secondary crash pairs, and 75 out of 708 crashes were identified as potential secondary crashes. Based on the preselected primary–secondary crash pairs, 17 secondary crashes were obtained with the proposed method and the results were compared with one of the commonly used methods, the speed contour plot method. Though the proposed method captured fewer secondary crashes, it did identify several secondary crashes that could not be observed with the speed contour plot method. The results showed the applicability of the method and the potential of crowdsourced Waze user reports in secondary crash identification.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2020

Fast shared-memory streaming multilevel graph partitioning

In this report we show that a fast parallel graph partitioner can benefit many applications by reducing data transfers. The online methods for partitioning graphs have to be fast and they often rely on simple one-pass streaming algorithms, while the offline methods for partitioning graphs contain more involved algorithms and the most successful methods in this category belong to the multilevel approaches. In this work, we assess the feasibility of using streaming graph partitioning algorithms within the multilevel framework. Our end goal is to come up with a fast parallel offline multilevel partitioner that can produce competitive cutsize quality. We rely on a simple but fast and flexible streaming algorithm throughout the entire multilevel framework. This streaming algorithm serves multiple purposes in the partitioning process: a clustering algorithm in the coarsening, an effective algorithm for the initial partitioning, and a fast refinement algorithm in the uncoarsening. Its simple nature also lends itself easily for parallelization. The experiments on various graphs show that our approach is on the average up to 5.1x faster than the multi-threaded MeTiS, which comes at the expense of only 2x worse cutsize.

97 MATHEMATICS AND COMPUTING↗