Seed-skewness algorithm for X-ray diffraction signal detection in time-resolved synchrotron Laue photocrystallography
Abstract not provided
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.
Abstract not provided
Next generation systems, such as edge devices, will need to provide efficient processing of machine learning (ML) algorithms along several metrics, including energy, performance, area, and latency. However, the quickly evolving field of ML makes it extremely difficult to generate accelerators able to support a wide variety of algorithms. At the same time, designing accelerators in hardware description languages (HDLs) by hand is hard and time consuming, and does not allow quick exploration of the design space. In this paper we present the Software Defined Accelerators From Learning Tools Environment (SODALITE), an automated open source high-level ML framework-to-verilog compiler targeting ML Application-Specific Integrated Circuits (ASICs) chiplets. The SODALITE approach will implement optimal designs by seamlessly combining custom components generated through high-level synthesis (HLS) with templated and fully tunable Intellectural Properties (IPs) and macros, integrated in an extendable resource library. Through a closed loop design space exploration engine, developers will be able to quickly explore their hardware designs along different dimensions.
This paper presents a hardware-in-the-loop cyber-physical system architecture design to monitor and control smart lights connected to the active distribution grid. The architecture uses Zigbee-based (IEEE 802.15.4) wireless sensor networks and publish-subscribe architecture to exchange monitoring and control signals between smart-light actuators (SLAs) and a smart-light central controller (SLCC). Each SLA integrated into a smart light consists of a Zigbee-based endpoint module to send and receive signals to and from the SLCC. The SLCC consists of a Zigbee-based coordinator module, which further exchanges the monitoring and control signals with the active distribution management system over the TCP/IP communication network. The monitoring signals from the SLAs include light status, brightness level, voltage, current, and power data, whereas, the control signals to the SLAs include light intensity, turn ON, turn OFF, standby, and default settings. We have used our existing hardware-in-the-loop (HIL) cyber-physical system (CPS) security SCADA testbed to process signals received from the SLCC and respond suitable control signals based on the smart light schedule requirements, system operation, and active distribution grid dynamic characteristics. We have integrated the proposed cyber-physical smart light control system (CPSLCS) testbed to our existing HIL CPS SCADA testbed. We use the integrated testbed to demonstrate the efficacy of the proposed algorithm by real-time performance and latency between the SLCC and SLAs. The experiments demonstrated significant results by 100% realtime performance and low latency while exchanging data between the SLCC and SLAs.
With the push to exascale, in situ visualization and analysis will continue to play an important role in high performance computing. Tightly coupling in situ visualization with simulations constrains resources for both, and these constraints force a complex balance of trade-offs. A performance model that provides an a priori answer for the cost of using an in situ approach for a given task would assist in managing the trade-offs between simulation and visualization resources. In this work, we present new statistical performance models, based on algorithmic complexity, that accurately predict the run-time cost of a set of representative rendering algorithms, an essential in situ visualization task. To train and validate the models, we conduct a performance study of an MPI+X rendering infrastructure used in situ with three HPC simulation applications. We then explore feasibility issues using the model for selected in situ rendering questions.
In this work, we present FastConv, a template-based code auto-generation open source library that can automatically generate high-performance deep learning convolution kernels of arbitrary matrices/tensors shapes. FastConv is based on the Winograd algorithm, which is reportedly the highest performing algorithm for the time-consuming convolution layers of convolutional neural networks. ARM CPUs cover a wide range designs and specifications, from embedded devices to HPC-grade CPUs. The leads to the dilemma of how to consistently optimize Winograd-based convolution solvers for convolution layers of different shapes. FastConv addresses this problem by using templates to auto-generate multiple shapes of tuned kernels variants suitable for skinny tall matrices. As a performance portable library, FastConv transparently searches for the best combination of kernel shapes, cache tiles, scheduling of loop orders, packing strategies, access patterns, and online/offline computations. Auto-tuning is used to search the parameter configuration space for the best performance for a given target architecture and problem size. The experiments with layer-wise evaluation on the VGG--16 model confirms a 1.25x performance gains is got by tuning the Winograd library. Integrated comparison results shows 1.02x to 1.40x, 1.14x to 2.17x, and 1.22x and 2.48x speedup is achieved over NNPACK, Arm NN, and FeatherCNN on the Kunpeng 920 beside few cases. Furthermore, problem size performance portability experiments with various convolution shapes shows that FastConv achieves 1.2x to 1.7x speedup and 2x to 22x speedup over NNPACK and ARM NN inference engine using Winograd on Kunpeng 920 . CPU performance portability evaluation on the VGG--16 show an average speedup over NNPACK of 1.42x, 1.21x, 1.26x, 1.37x, 2.26x, and 11.02x is observed on Kunpeng 920, Snapdragon 835, 855, 888, Apple M1, and AWS Graviton2, respectively.
Historically, centrally computed algorithms have been the primary means of power system optimization and control. With increasing penetrations of distributed energy resources requiring optimization and control of power systems with many controllable devices, distributed algorithms have been the subject of significant research interest. Here, this paper surveys the literature of distributed algorithms with applications to optimization and control of power systems. In particular, this paper reviews distributed algorithms for offline solution of optimal power flow (OPF) problems as well as online algorithms for real-time solution of OPF, optimal frequency control, optimal voltage control, and optimal wide-area control problems.
Modern power systems (MPSs), including microgrids (MGs), are increasingly incorporating multiple renewable energy sources (RESs) such as wind and solar power, as well as battery storage and controllable loads. While environmentally beneficial, these sources pose challenges for control and management due to their intermittent and stochastic nature, especially in maintaining frequency stability with multiple interconnected generators of varying capacities. Traditional droop control methods are effective in systems with generators that are dispatchable and have fixed generation capacities, but they fall short when applied to systems with RESs, where generation capacities are dynamic and affected by unpredictable environmental conditions. To address these challenges, this paper introduces a novel stochastic adaptive droop control (SADC) method for load frequency control (LFC). The proposed method adapts droop coefficients in real time, based on the measured stochastic data of power generation capacities, enabling more effective frequency regulation in systems with variable and intermittent power generation. Unlike traditional adaptive control methods, which assume constant or slowly-varying system parameters, this approach accounts for stochastic processes by modeling them as Markov chains, enabling robust performance under highly dynamic and unpredictable conditions. The key contributions of this work include the development of real-time droop coefficient adaptation algorithms, derivation of their stability and convergence properties, and the demonstration of the advantages of the method through simulations. Case studies highlight the improved performance of frequency regulation, particularly in addressing the impact of stochastic weather conditions and the benefits of reducing dependence on battery reserves in dealing with intermittency of RESs. Finally, this paper provides a comprehensive analysis of the theoretical foundations of the method, as well as practical implementation insights for future power systems with high penetration of RESs.
Here, we present a novel online system identification methodology for monitoring the performance of power systems. This methodology was demonstrated in a gas turbine recuperated power plant designed for a hybrid configuration. A 120-kW Garrett microturbine modified to test dynamic control strategies for hybrid power systems designed at the National Energy Technology Laboratory (NETL) was used to implement and validate this online system identification methodology. The main component of this methodology consists of an empirical transfer function model implemented in parallel to the turbine speed operation and the fuel control valve, which can monitor the process response of the gas turbine system while it is operating. During fully closed-loop operations or automated control, the output of the controller, fuel valve position, and the turbine speed measurements were fed for a given period of time to a recursive algorithm that determined the transfer function parameters during the nominal condition. After the new parameters were calculated, they were fed into the transfer function model for online prediction. The turbine speed measurement was compared against the transfer function prediction, and a control logic was implemented to capture when the system operated at nominal or abnormal conditions. To validate the ability to detect abnormal conditions during dynamic operations, drifting in the performance of the gas turbine system was evaluated. A leak in the turbomachinery working fluid was emulated by bleeding 10% of the airflow from the compressor discharge to the atmosphere, and electrical load steps were performed before and after the leak. This tool could detect the leak 7 s after it had occurred, which accounted for a fuel flow increase of approximately 15.8% to maintain the same load and constant turbine speed operations.
Purpose: The main goal of this paper is to reconstruct a distribution of cell survival fractions from tumor-volume variation for a heterogeneous group of head and neck cancer patients and compare this distribution to the data from predictive assays. Methods: To characterize the tumor-volume variation during radiation therapy treatment, the authors use a two-level tumor-volume model of cell population that separates the entire tumor cell population into two subpopulations of viable cells and lethally damaged cells. This parameterized radiobiological model is integrated with a least squares objective function and a simulated annealing optimization algorithm to describe time-dependent tumor-volume variation rates in individual patients. Several constraints have been used in the optimization problem because tumor-volume variation during radiotherapy is described by a sum of exponentials; therefore, the problem of accurately fitting a model to measured data is ill-posed. The model was applied to measured tumor-volume variation curves from a clinical study on tumor-volume variation during radiotherapy for 14 head and neck cancer patients in which an integrated CT/linear particle accelerator (LINAC) system was used for tumor-volume measurements. Results: The two-level cell population tumor-volume modeling is capable of describing tumor-volume variation throughout the entire treatment for 11 of the 14 patients. For three patients, the tumor-volume variation was described only during the initial part of treatment, a fact that may be related to the neglected hypoxia in the two-level approximation. The predicted probability density distribution for the survival fractions agrees with the data obtained using in vitro studies with predictive assays. The mean value 0.35 of survival fraction obtained in this study is larger than the value 0.32 from in vitro studies, which could be expected because of greater repair in vivo. The mean half-life obtained in this study for the head-and-neck squamous cell carcinoma (SCC) is equal to 3.8 mean potential doubling times, which agrees with 4.0 mean potential doubling times obtained previously for lung SCC. Conclusions: The distribution of cell survival fractions obtained in this study support the hypothesis that the tumor-volume variation during radiotherapy treatment for head and neck cancer can be described by the two-level cell population tumor-volume model. This model can be used for in vivo evaluation of patient-specific radiobiological parameters that are needed for tumor-control probability evaluation.
Purpose: Prostate adenocarcinoma is the most common noncutaneous malignancy in American men with over 200 000 new cases diagnosed each year. Prostate interventional therapy, such as cryotherapy and brachytherapy, is an effective treatment for prostate cancer. Its success relies on the correct needle implant position. This paper proposes a robust and efficient needle segmentation method, which acts as an aid to localize the needle in three-dimensional (3D) transrectal ultrasound (TRUS) guided prostate therapy. Methods: The procedure of locating the needle in a 3D TRUS image is a three-step process. First, the original 3D ultrasound image containing a needle is cropped; the cropped image is then converted to a binary format based on its histogram. Second, a 3D Hough transform based needle segmentation method is applied to the 3D binary image in order to locate the needle axis. The position of the needle endpoint is finally determined by an optimal threshold based analysis of the intensity probability distribution. The overall efficiency is improved through implementing a coarse-fine searching strategy. The proposed method was validated in tissue-mimicking agar phantoms, chicken breast phantoms, and 3D TRUS patient images from prostate brachytherapy and cryotherapy procedures by comparison to the manual segmentation. The robustness of the proposed approach was tested by means of varying parameters such as needle insertion angle, needle insertion length, binarization threshold level, and cropping size. Results: The validation results indicate that the proposed Hough transform based method is accurate and robust, with an achieved endpoint localization accuracy of 0.5 mm for agar phantom images, 0.7 mm for chicken breast phantom images, and 1 mm for in vivo patient cryotherapy and brachytherapy images. The mean execution time of needle segmentation algorithm was 2 s for a 3D TRUS image with size of 264 Multiplication-Sign 376 Multiplication-Sign 630 voxels. Conclusions: The proposed needle segmentation algorithm is accurate, robust, and suitable for 3D TRUS guided prostate transperineal therapy.
Purpose: Monte Carlo (MC) simulation is an important tool to solve radiotherapy and medical imaging problems. Low computational efficiency hinders its wide applications. Conventionally, MC is performed in a particle-by -particle fashion. The lack of control on particle trajectory is a main cause of low efficiency in some applications. Take cone beam CT (CBCT) projection simulation as an example, significant amount of computations were wasted on transporting photons that do not reach the detector. To solve this problem, we propose an innovative MC simulation scheme with a path-by-path sampling method. Methods: Consider a photon path starting at the x-ray source. After going through a set of interactions, it ends at the detector. In the proposed scheme, we sampled an entire photon path each time. Metropolis-Hasting algorithm was employed to accept/reject a sampled path based on a calculated acceptance probability, in order to maintain correct relative probabilities among different paths, which are governed by photon transport physics. We developed a package gMMC on GPU with this new scheme implemented. The performance of gMMC was tested in a sample problem of CBCT projection simulation for a homogeneous object. The results were compared to those obtained using gMCDRR, a GPU-based MC tool with the conventional particle-by-particle simulation scheme. Results: Calculated scattered photon signals in gMMC agreed with those from gMCDRR with a relative difference of 3%. It took 3.1 hr. for gMCDRR to simulate 7.8e11 photons and 246.5 sec for gMMC to simulate 1.4e10 paths. Under this setting, both results attained the same ∼2% statistical uncertainty. Hence, a speed-up factor of ∼45.3 was achieved by this new path-by-path simulation scheme, where all the computations were spent on those photons contributing to the detector signal. Conclusion: We innovatively proposed a novel path-by-path simulation scheme that enabled a significant efficiency enhancement for MC particle transport simulations.
High level abstractions in Python that can utilize computing hardware well seem to be an attractive option for writing data reduction and analysis tasks. In this paper, we explore the features available in Python which are useful and efficient for end user analysis in High Energy Physics (HEP). A typical vertical slice of an HEP data analysis is somewhat fragmented: the state of the reduction/analysis process must be saved at certain stages to allow for selective reprocessing of only parts of a generally time-consuming workflow. Also, algorithms tend to to be modular because of the heterogeneous nature of most detectors and the need to analyze different parts of the detector separately before combining the information. This fragmentation causes difficulties for interactive data analysis, and as data sets increase in size and complexity (O10 TiB for a “small” neutrino experiment to the O10 PiB currently held by the CMS experiment at the LHC), data analysis methods traditional to the field must evolve to make optimum use of emerging HPC technologies and platforms. Mainstream big data tools, while suggesting a direction in terms of what can be done if an entire data set can be available across a system and analysed with high-level programming abstractions, are not designed with either scientific computing generally, or modern HPC platform features in particular, such as data caching levels, in mind. Our example HPC use case is a search for a new elementary particle which might explain the phenomenon known as “Dark Matter”. Here, using data from the CMS detector, we will use HDF5 as our input data format, and MPI with Python to implement our use case.
Despite the apparent cross-disciplinary interactions among scientific fields, a formal description of their evolution is lacking. Here we describe a novel approach to study the dynamics and evolution of scientific fields using a network-based analysis. We build an idea network consisting of American Physical Society Physics and Astronomy Classification Scheme (PACS) numbers as nodes representing scientific concepts. Two PACS numbers are linked if there exist publications that reference them simultaneously. We locate scientific fields using a community finding algorithm, and describe the time evolution of these fields over the course of 1985–2006. The communities we identify map to known scientific fields, and their age depends on their size and activity. We expect our approach to quantifying the evolution of ideas to be relevant for making predictions about the future of science and thus help to guide its development.
Traditionally, the kinetics of DNA repair have been estimated using immunocytochemistry by labeling proteins involved in the DNA damage response (DDR) with fluorescent markers in a fixed cell assay. However, detailed knowledge of DDR dynamics across multiple cell generations cannot be obtained using a limited number of fixed cell time-points. Here we report on the dynamics of 53BP1 radiation induced foci (RIF) across multiple cell generations using live cell imaging of non-malignant human mammary epithelial cells (MCF10A) expressing histone H2B-GFP and the DNA repair protein 53BP1-mCherry. Using automatic extraction of RIF imaging features and linear programming techniques, we were able to characterize detailed RIF kinetics for 24 hours before and 24 hours after exposure to low and high doses of ionizing radiation. High-content-analysis at the single cell level over hundreds of cells allows us to quantify precisely the dose dependence of 53BP1 protein production, RIF nuclear localization and RIF movement after exposure to X-ray. Using elastic registration techniques based on the nuclear pattern of individual cells, we could describe the motion of individual RIF precisely within the nucleus. We show that DNA repair occurs in a limited number of large domains, within which multiple small RIFs form, merge and/or resolve with random motion following normal diffusion law. Large foci formation is shown to be mainly happening through the merging of smaller RIF rather than through growth of an individual focus. We estimate repair domain sizes of 7.5 to 11 µm 2 with a maximum number of ~15 domains per MCF10A cell. This work also highlights DDR which are specific to doses larger than 1 Gy such as rapid 53BP1 protein increase in the nucleus and foci diffusion rates that are significantly faster than for spontaneous foci movement. We hypothesize that RIF merging reflects a "stressed" DNA repair process that has been taken outside physiological conditions when too many DSB occur at once. High doses of ionizing radiation lead to RIF merging into repair domains which in turn increases DSB proximity and misrepair. Furthermore, such finding may therefore be critical to explain the supralinear dose dependence for chromosomal rearrangement and cell death measured after exposure to ionizing radiation.
A current area of research interest in national security is to effectively and efficiently determine the contents of the many shipping containers that enter ports in the United States. This interest comes as a result of the 9/11 Commission Act passed by Congress in 2007 that requires 100% of inbound cargo to be scanned by 2012. It appears that this requirement will be achieved by 2012, but as of February of 2009 eighty percent of the 11.5 million inbound cargo containers were being scanned. The systems used today in all major U.S. ports to determine the presence of radioactive material within cargo containers are Radiation Portal Monitors (RPM). These devices generally exist in the form of a gate or series of gates that the containers can be driven through and scanned. The monitors are effective for determining the presence of radiation, but offer little more information about the particular source. This simple pass-fail system leads to many false alarms as many everyday items emit radiation including smoke detectors due to the Americium-241 source contained inside, bananas, milk, cocoa powder and lean beef due to the trace amounts of Potassium-40, and fire brick and kitty litter due to their high clay content which often contains traces of uranium and thorium. In addition, if an illuminating source is imposed on the boundary of the container, the contents of the container may become activated. These materials include steel, aluminum and many agricultural products. Current portal monitors also have not proven to be that effective at identifying natural or highly enriched uranium (HEU). In fact, the best available Advanced Spectroscopic Portal Monitors (ASP) are only capable of identifying bare HEU 70-88% of the time and masked HEU and depleted uranium (DU) only 53 percent of the time. Therefore, a better algorithm that uses more information collected from better detectors about the specific material distribution within the container is desired. The work reported here explores the inverse problem of optical tomography applied to heterogeneous domains. The neutral particle transport equation was used as the forward model for how neutral particles stream through and interact within these heterogeneous domains. A constrained optimization technique that uses Newtons method served as the basis of the inverse problem. Optical tomography aims at reconstructing the material properties using (a) illuminating sources and (b) detector readings. However, accurate simulations for radiation transport require that the particle (gamma and/or neutron) energy be appropriate discretize in the multigroup approximation. This, in turns, yields optical tomography problems where the number of unknowns grows (1) about quadratically with respect to the number of energy groups, G, (notably to reconstruct the scattering matrix) and (2) linearly with respect to the number of unknown material regions. As pointed out, a promising approach could rely on algorithms to appropriately select a material type per material zone rather than G2 values. This approach, though promising, still requires further investigation: (a) when switching from cross-section values unknowns to material type indices (discrete integer unknowns), integer programming techniques are needed since derivative information is no longer available; and (b) the issue of selecting the initial material zoning remains. The work reported here proposes an approach to solve the latter item, whereby a material zoning is proposed using one-group or few-groups transport approximations. The capabilities and limitations of the presented method were explored; they are briefly summarized next and later described in fuller details in the Appendices. The major factors that influenced the ability of the optimization method to reconstruct the cross sections of these domains included the locations of the sources used to illuminate the domains, the number of separate experiments used in the reconstruction, the locations where measurements were collected, the optical thickness of the domain, the amount of signal noise and signal bias applied to the measurements and the initial guess for the cross section distribution. All of these factors were explored for problems with and without scattering. Increasing the number of source and measurement locations and experiments generally was more successful at reconstructing optically thicker domains while producing less error in the image. The maximum optical thickness that could be reconstructed with this method was ten mean free paths for pure absorber and two mean free paths for scattering problems. Applying signal noise and signal bias to the measured fluxes produced more error in the produced image. Generally, Newtons method was more successful at reconstructing domains from an initial guess for the cross sections that was greater in magnitude than their true values than from an initial guess that was lower in magnitude.
Here we report on MIT’s THM work in Q4 2016 and Q1 2017. The goal of this project is to design, construct and execute tests of flow boiling critical heat flux (CHF) at high-pressure using high-resolution and high-speed video and infrared (IR) thermometry, to generate unique data to inform the development of and validate mechanistic boiling heat transfer and CHF models. In FY2016, a new test section was designed and fabricated. Data was collected at atmospheric conditions at 10, 25 and 50 K subcoolings, and three mass fluxes, i.e. 500, 750 and 1000 kg/m2/s. Starting in Q4 2016 and continuing forward, new post-processing techniques have been developed to analyze the data collected. These new algorithms analyze the time-dependent temperature and heat flux distributions to calculate nucleation site density, nucleation frequency, growth and wait time, dry area fraction, and the complete heat flux partitioning. In Q1 2017 a new flow boiling loop was designed and constructed to support flow boiling tests up 10 bar pressure and 180 °C. Initial shakedown and testing has been completed. The flow loop and test section are now ready to begin high-pressure flow boiling testing.
This report summarizes experimental radiation detector, contextual sensor, weather, and global positioning system (GPS) data collected to inform and validate a comprehensive, operational radiation transport modeling framework to evaluate radiation detector system and algorithm performance. This framework will be used to study the influence of systematic effects (such as geometry, background activity, background variability, environmental shielding, etc.) on detector responses and algorithm performance using synthetic time series data. This work consists of performing data collection campaigns at a canonical, controlled environment for complete radiological characterization to help construct and benchmark a high-fidelity model with quantified system geometries, detector response functions, and source terms for background and threat objects. This data also provides an archival, benchmark dataset that can be used by the radiation detection community. The data reported here spans four data collection campaigns conducted between May 2015 and September 2016.
This project is focused on early stage research and development of numerical methods and models to improve advanced engine combustion concepts and systems. The current focus is on development of new mathematics and algorithms to reduce the time to solution for advanced combustion engine design using detailed fuel chemistry. The research is prioritized towards the most time-consuming workflow bottlenecks (computer and human) and accuracy gaps that slow ACS program members. Zero-RK, the fast and accurate chemical kinetics solver software developed in this project, is central to the research efforts and continues to be developed to address the current and emerging needs of the engine designers, engine modelers and fuel mechanism developers.