Search NASA⌕ Search

SEARCH · Search NASA

Results for “sequential updates”

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 91 records · Page 5

On the Convergence of Overlapping Schwarz Decomposition for Nonlinear Optimal Control

Here, we study the convergence properties of an overlapping Schwarz decomposition algorithm for solving nonlinear optimal control problems (OCPs). The algorithm decomposes the time domain into a set of overlapping subdomains, and solves all subproblems defined over subdomains in parallel. The convergence is attained by updating primal-dual information at the boundaries of overlapping subdomains. We show that the algorithm exhibits local linear convergence, and that the convergence rate improves exponentially with the overlap size. We also establish global convergence results for a general quadratic programming, which enables the application of the Schwarz scheme inside second-order optimization algorithms (e.g., sequential quadratic programming). The theoretical foundation of our convergence analysis is a sensitivity result of nonlinear OCPs, which we call "exponential decay of sensitivity" (EDS). Intuitively, EDS states that the impact of perturbations at domain boundaries (i.e., initial and terminal time) on the solution decays exponentially as one moves into the domain. Here, we expand a previous analysis available in the literature by showing that EDS holds for both primal and dual solutions of nonlinear OCPs, under uniform second-order sufficient condition, controllability condition, and boundedness condition. We conduct experiments with a quadrotor motion planning problem and a partial differential equations (PDE) control problem to validate our theory, and show that the approach is significantly more efficient than alternating direction method of multipliers and as efficient as the centralized interior-point solver.

42 ENGINEERING↗

Layer-Parallel Training of Deep Residual Neural Networks

Residual neural networks (ResNets) are a promising class of deep neural networks that have shown excellent performance for a number of learning tasks, e.g., image classification and recognition. Mathematically, ResNet architectures can be interpreted as forward Euler discretizations of a nonlinear initial value problem whose time-dependent control variables represent the weights of the neural network. Hence, training a ResNet can be cast as an optimal control problem of the associated dynamical system. For similar time-dependent optimal control problems arising in engineering applications, parallel-in-time methods have shown notable improvements in scalability. This paper demonstrates the use of those techniques for efficient and effective training of ResNets. The proposed algorithms replace the classical (sequential) forward and backward propagation through the network layers with a parallel nonlinear multigrid iteration applied to the layer domain. This adds a new dimension of parallelism across layers that is attractive when training very deep networks. From this basic idea, we derive multiple layer-parallel methods. The most efficient version employs a simultaneous optimization approach where updates to the network parameters are based on inexact gradient information in order to speed up the training process. Finally, using numerical examples from supervised classification, we demonstrate that the new approach achieves a training performance similar to that of traditional methods, but enables layer-parallelism and thus provides speedup over layer-serial methods through greater concurrency.

97 MATHEMATICS AND COMPUTING↗

The HITRAN2016 molecular spectroscopic database

This paper describes the contents of the 2016 edition of the HITRAN molecular spectroscopic compilation. The new edition replaces the previous HITRAN edition of 2012 and its updates during the intervening years. The HITRAN molecular absorption compilation is comprised of five major components: the traditional line-by-line spectroscopic parameters required for high-resolution radiative-transfer codes, infrared absorption cross-sections for molecules not yet amenable to representation in a line-by-line form, collision-induced absorption data, aerosol indices of refraction, and general tables such as partition sums that apply globally to the data. The new HITRAN is greatly extended in terms of accuracy, spectral coverage, additional absorption phenomena, added line-shape formalisms, and validity. Moreover, molecules, isotopologues, and perturbing gases have been added that address the issues of atmospheres beyond the Earth. Of considerable note, experimental IR cross-sections for almost 200 additional significant molecules have been added to the database.

Gordon, I. E.↗

Highly scalable distributed-memory sparse triangular solution algorithms.

This paper presents a highly efficient distributed-memory parallel sparse triangular solver. The triangular solution phase is often performed following factorization phase in the sparse linear solvers and has become increasingly computa­tionally expensive for direct solvers with many right hand sides (RHSs) or preconditioned iterative solvers. However, the low arithmetic intensity and sequential nature of the triangular solve algorithm pose performance challenges for its large-scale distributed-memory parallelization. In this work, we propose several strategies to enhance scalability of an algorithm with 2D block cyclic process layout. First, an asynchronous binary-tree-based communication scheme implemented via non-blocking MPI functions is leveraged to broadcast partial solutions and reduce partial updates among a subset of processes for each block column and row of the triangular matrix, respectively. This scheme reduces message latency, improves communication load balance and significantly accelerates asynchronous execution of the triangular solve. In addition, efficient BLAS operations and threading implementations are exploited to accelerate local computations and further reduce process idle time. The proposed strategies are implemented in SuperLU_DIST and numerical experiments show up to 4.4x improvement with one right-hand side and up to 6.1x improvement with 50 right- hand sides on 4096 processes, compared to the current release. This is the first time that sparse triangular solution is demonstrated strong scaling on more than 4000 cores.

Liu, Yang↗

A Monte Carlo Rectangle Packing Algorithm for Identifying Likely Spatial Distributions of Final Closure Cap Subsidence in the E-Area Low-Level Waste Facility

A Monte Carlo rectangle packing algorithm, summarized in Figure ES-1, has been developed to provide insights into the likely spatial distribution of subsided areas or holes in the final closure cap for the E-Area Low-Level Waste Facility (ELLWF) resulting from the placement of non-crushable waste containers within slit and engineered trench disposal units. Using historical records provided by Solid Waste Management as an indicator of future waste streams, a probability distribution of unique waste containers was constructed, and Monte Carlo sampling was used to sequentially select waste containers for placement in a rectangular region bounded by the disposal unit footprint. Within this Monte Carlo approach, the probability of any container type being selected is weighted by the fractional abundance relative to all other container types. For each Monte Carlo realization, the frequency of non-crushable waste containers appearing at each location on a spatially discretized mesh is updated. After many realizations, the location(s) with the highest frequencies of non-crushable waste containers are regarded as the most likely locations to experience subsidence (i.e., for a hole to form). Utilizing this approach has enabled the down-selection of a computationally intractable set of subsidence hole combinations to be explored in three-dimensional finite element groundwater flow and contaminant transport models. Such an approach can be applied to a number of different packaging problems.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Cybersecurity Enhancement for Multi-Infeed High-Voltage DC Systems

Composed of multiple two-terminal high-voltage DC (HVDC) transmission systems, a multi-infeed HVDC (MIDC) system exchanges massive power among multiple asynchronous AC systems. However, as an intrinsically cyber-physical system, an MIDC system could suffer from cyber-attacks, leading to massive power mismatches in multiple AC systems, and resulting in catastrophic consequences. Since the sequential responses of an MIDC system and interconnected AC systems are in different timescales, this paper first establishes a two-timescale model to evaluate the sequential impacts caused by cyber-attacks. Then, an event-triggered cyber-defense strategy is proposed to enhance the cybersecurity of an MIDC system by mitigating multiple non-simultaneous cyber-attacks. Whenever new cyber-attack events occur, the proposed cyber-defense strategy, which is mathematically modeled as a mixed-integer quadratic programming problem, is executed on-line and updated in an event-triggered manner. Here, simulation results on an MIDC system demonstrate that the low-cost and almost blind cyber-attacks can cause severe frequency deviations, and the proposed strategy can mitigate multi-cyber-attacks effectively.

24 POWER TRANSMISSION AND DISTRIBUTION↗

High-Dose Hypofractionated Proton Beam Radiation Therapy Is Safe and Effective for Central and Peripheral Early-Stage Non-Small Cell Lung Cancer: Results of a 12-Year Experience at Loma Linda University Medical Center

Purpose: We update our previous reports on the use of hypofractionated proton beam radiation therapy for early-stage lung cancer patients. Methods and Materials: Eligible subjects had biopsy-proven non-small cell carcinoma of the lung and were medically inoperable or refused surgery. Clinical workup required staging of T1 or T2, N0, M0. Subjects received hypofractionated proton beam therapy to the primary tumor only. The dose delivered was sequentially escalated from 51 to 60 Gy, then to 70 Gy in 10 fractions over 2 weeks. Endpoints included toxicity, pulmonary function, overall survival (OS), disease-specific survival (DSS), and local control (LC). Results: One hundred eleven subjects were analyzed for treatment outcomes. The patient population had the following average characteristics; age 73.2 years, tumor size 3.6 cm, and 1.33 L forced expiratory volume in 1 second. The entire group showed improved OS with increasing dose level (51, 60, and 70 Gy) with a 4-year OS of 18%, 32%, and 51%, respectively (P=.006). Peripheral T1 tumors exhibited LC of 96%, DSS of 88%, and OS of 60% at 4 years. Patients with T2 tumors showed a trend toward improved LC and survival with the 70-Gy dose level. On multivariate analysis, larger tumor size was strongly associated with increased local recurrence and decreased survival. Central versus peripheral location did not correlate with any outcome measures. Clinical radiation pneumonitis was not found to be a significant complication, and no patient required steroid therapy after treatment for radiation pneumonitis. Pulmonary function was well maintained 1 year after treatment. Conclusions: High-dose hypofractionated proton therapy achieves excellent outcomes for lung carcinomas that are peripherally or centrally located. The 70-Gy regimen has been adopted as standard therapy for T1 tumors at our institution. Larger T2 tumors show a trend toward improved outcomes with higher doses, suggesting that better results could be seen with intensified treatment.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Sequential Design of Experiments to Maximize Learning from Carbon Capture Pilot Plant Testing

Pilot plant test campaigns can be expensive and time-consuming. Therefore, it is of interest to maximize the amount of learning and the efficiency of the test campaign given the limited number of experiments that can be conducted. This work investigates the use of sequential design of experiments (SDOE) to overcome these challenges by demonstrating its usefulness for a recent solvent-based CO2 capture plant test campaign. Unlike traditional design of experiments methods, SDOE regularly uses information from ongoing experiments to determine the optimum locations in the design space for subsequent runs within the same experiment. However, there are challenges that need to be addressed, including reducing the high computational burden to efficiently update the model, and the need to incorporate the methodology into a computational tool. We address these challenges by applying SDOE in combination with a software tool, the Framework for Optimization, Quantification of Uncertainty and Surrogates (FOQUS) (Miller et al., 2014a, 2016, 2017). The results of applying SDOE on a pilot plant test campaign for CO2 capture suggests that relative to traditional design of experiments methods, SDOE can more effectively reduce the uncertainty of the model, thus decreasing technical risk. Future work includes integrating SDOE into FOQUS and using SDOE to support additional large-scale pilot plant test campaigns.

20 FOSSIL-FUELED POWER PLANTS↗

Data assimilation in operator algebras

We develop an algebraic framework for sequential data assimilation of partially observed dynamical systems. In this framework, Bayesian data assimilation is embedded in a nonabelian operator algebra, which provides a representation of observables by multiplication operators and probability densities by density operators (quantum states). In the algebraic approach, the forecast step of data assimilation is represented by a quantum operation induced by the Koopman operator of the dynamical system. Moreover, the analysis step is described by a quantum effect, which generalizes the Bayesian observational update rule. Projecting this formulation to finite-dimensional matrix algebras leads to computational schemes that are i) automatically positivity-preserving and ii) amenable to consistent data-driven approximation using kernel methods for machine learning. Moreover, these methods are natural candidates for implementation on quantum computers. Applications to the Lorenz 96 multiscale system and the El Niño Southern Oscillation in a climate model show promising results in terms of forecast skill and uncertainty quantification.

97 MATHEMATICS AND COMPUTING↗

CCSI Toolset 3.21 Release

CCSI Toolset 3.21 Release Highlights Parallelization support was added for Sequential Design of Experiments (SDOE) computations using Dask (preliminary). Input type dependent ordering capability was added to the SDOE module. With this implementation the user can specify the level of difficulty to change an input (Easy or Hard) and FOQUS will generate the appropriate ordered design depending on the input difficulty combination. Python version support was extended. FOQUS is now compatible with Python 3.8 through 3.12. Platforms used for automated testing were expanded to include macOS ARM (Apple Silicon). Updates to the FOQUS documentation to include information on how to set paths for SimSinter and TurbineLite. Turbine configuration section was added to Debugging Documentation.

AS↗

Optimal Power Flow of Radial Networks and Its Variations: A Sequential Convex Optimization Approach

This paper proposes a sequential convex optimization method to solve broader classes of optimal power flow (OPF) problems over radial networks. The non-convex branch power flow equation is decomposed as a second-order cone inequality and a non-convex constraint involving the difference of two convex functions. Provided with an initial solution offered by an inexact second-order cone programming relaxation model, this approach solves a sequence of convexified penalization problems, where concave terms are approximated by linear functions and updated in each iteration. It could recover a feasible power flow solution, which usually appears to be very close, if not equal, to the global optimal one. Two variations of the OPF problem, in which non-cost related objectives are optimized subject to power flow constraints and the convex relaxation is generally inexact, are elaborated in detail. One is the maximum loadability problem, which is formulated as a special OPF problem that seeks the maximal distance to the boundary of power flow insolvability. The proposed method is shown to outperform commercial nonlinear solvers in terms of robustness and efficiency. The other is the bi-objective OPF problem. A non-parametric scalarization model is suggested, and is further reformulated as an extended OPF problem by convexifying the objective function. It provides a single trade-off solution without any subjective preference. The proposed computation framework also helps retrieve the Pareto front of the bi-objective OPF via the e-constraint method or the normal boundary intersection method. This paper also discusses extensions for OPF problems over meshed networks based on the semidefinite programming relaxation method.

bi-objective optimization↗

Accelerating solidification process simulation for large-sized system of liquid metal atoms using GPU with CUDA

Molecular dynamics simulation is a powerful tool to simulate and analyze complex physical processes and phenomena at atomic characteristic for predicting the natural time-evolution of a system of atoms. Precise simulation of physical processes has strong requirements both in the simulation size and computing timescale. Therefore, finding available computing resources is crucial to accelerate computation. However, a tremendous computational resource (GPGPU) are recently being utilized for general purpose computing due to its high performance of floating-point arithmetic operation, wide memory bandwidth and enhanced programmability. As for the most time-consuming component in MD simulation calculation during the case of studying liquid metal solidification processes, this paper presents a fine-grained spatial decomposition method to accelerate the computation of update of neighbor lists and interaction force calculation by take advantage of modern graphics processors units (GPU), enlarging the scale of the simulation system to a simulation system involving 10 000 000 atoms. In addition, a number of evaluations and tests, ranging from executions on different precision enabled-CUDA versions, over various types of GPU (NVIDIA 480GTX, 580GTX and M2050) to CPU clusters with different number of CPU cores are discussed. The experimental results demonstrate that GPU-based calculations are typically 9∼11 times faster than the corresponding sequential execution and approximately 1.5∼2 times faster than 16 CPU cores clusters implementations. On the basis of the simulated results, the comparisons between the theoretical results and the experimental ones are executed, and the good agreement between the two and more complete and larger cluster structures in the actual macroscopic materials are observed. Moreover, different nucleation and evolution mechanism of nano-clusters and nano-crystals formed in the processes of metal solidification is observed with large-sized system.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Switching Device-Cognizant Sequential Distribution System Restoration

This paper presents an optimization framework for sequential reconfiguration using an assortment of switching devices and repair process in distribution system restoration. Compared to existing studies, this paper considers types, capabilities and operational limits of different switching devices, making it applicable in practice. We develop a novel multi-phase method to find the optimal sequential operation of various switching devices and repair faulted areas. We consider circuit breakers, reclosers, sectionalizers, load breaker switches, and fuses. The switching operation problem is decomposed into two mixed-integer linear programming (MILP) subproblems. The first subproblem determines the optimal network topology and estimates the number of steps to reach that topology, while the second subproblem generates a sequence of switching operations to coordinate the switches. For repairing the faults, we design an MILP model that dispatches repair crews to clear faults and replace melted fuses. After clearing a fault, we update the topology of the network by generating a new sequence of switching operations, and the process continues until all faults are cleared. To improve the computational efficiency, a network reduction algorithm is developed to group line sections, such that only switchable sections are present in the reduced network. The proposed method is validated on the IEEE 123-bus and 8500-bus systems.

distribution system↗

Calibration and Uncertainty Estimation Using the Ensemble Kalman Filter with a Large Subsurface Flow and Transport Model - 20321

At routinely monitored groundwater contamination sites, periodically measured environmental conditions such as groundwater levels and contaminant concentrations are used to inform and confirm a conceptual site model (CSM) and guide the development and calibration of a numerical groundwater flow and transport model. The calibration of groundwater flow and transport models after each measurement (sampling) event can illuminate deficiencies in a CSM, identify areas where additional monitoring is warranted, and predict the behavior of the system to guide decision making. However, manual and automated (e.g. PEST) model calibration tools can be time-consuming and computationally expensive to implement after each sampling event. Perhaps as a result, such calibration tools generally utilize all available monitoring data simultaneously rather than sequentially assimilating monitoring data one sampling event at a time as the results from sampling become available. A more real-time data assimilation approach may reduce parameter uncertainty, quantify the value of additional monitoring data, and produce a usable model more quickly and with less effort. To mitigate the potential time-consuming aspects of manual and widely applied automated calibration techniques, a data assimilation algorithm called the ensemble Kalman filter (EnKF) was evaluated as a relatively efficient method of model calibration and uncertainty assessment via the sequential integration of monitoring data into a model. The EnKF was able to successfully and efficiently assimilate monitoring and modeling data to calibrate a complex flow and transport model at a real-world site with significant subsurface heterogeneity, uncertainty, and 12 years of monitoring data (over 4,000 individual measurements of groundwater levels and over 2,500 measurements of contaminant concentrations). Starting with an uncalibrated model data from annual sampling events were sequentially assimilated, and the resultant predication errors and estimated parameter uncertainties were tracked. After all monitoring data were assimilated, both flow and transport residuals at the end of the EnKF process were comparable to those produced via a concurrent PEST calibration effort but required fewer model simulations. Both uncertainty and prediction errors decreased over time. In a real-time application, the adequacy of the model could be assessed after each sampling event. The benefits of such a real-time approach to utilizing monitoring data include reduced costs (in the form of model updates or site characterization efforts), early flagging of possible errors in the CSM, and a reduced risk of overfitting and corresponding increased confidence in model predictions. This tool may be particularly useful compared to other calibration techniques (e.g. manual, PEST) when model runtimes are long, calibration parameters are many, or parameter uncertainty is large. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Systems and methods for enhanced sequential power system model parameter estimation

A system for enhanced sequential power system model calibration is provided. The system is programmed to store a model of a device. The model includes a plurality of parameters. The system is also programmed to receive a plurality of events associated with the device, receive a first set of calibration values for the plurality of parameters, generate a plurality of sets of calibration values for the plurality of parameters, for each of the plurality of sets of calibration values, analyze a first event of the plurality of events using a corresponding set of calibration values to generate a plurality of updated sets of calibration values, analyze the plurality of updated sets of calibration values to determine a current updated set of calibration values, and update the model to include the current updated set of calibration values.

Wang, Honggang↗

Blind identification of full-field vibration modes of output-only structures from uniformly-sampled, possibly temporally-aliased (sub-Nyquist), video measurements

Enhancing the spatial and temporal resolution of vibration measurements and modal analysis could significantly benefit dynamic modelling, analysis, and health monitoring of structures. For example, spatially high-density mode shapes are critical for accurate vibration-based damage localization. In experimental or operational modal analysis, higher (frequency) modes, which may be outside the frequency range of the measurement, contain local structural features that can improve damage localization as well as the construction and updating of the modal-based dynamic model of the structure. In general, the resolution of vibration measurements can be increased by enhanced hardware. Traditional vibration measurement sensors such as accelerometers have high-frequency sampling capacity; however, they are discrete point-wise sensors only providing sparse, low spatial sensing resolution measurements, while dense deployment to achieve high spatial resolution is expensive and results in the mass-loading effect and modification of structure's surface. Non-contact measurement methods such as scanning laser vibrometers provide high spatial and temporal resolution sensing capacity; however, they make measurements sequentially that requires considerable acquisition time. As an alternative non-contact method, digital video cameras are relatively low-cost, agile, and provide high spatial resolution, simultaneous, measurements. Combined with vision based algorithms (e.g., image correlation or template matching, optical flow, etc.), video camera based measurements have been successfully used for experimental and operational vibration measurement and subsequent modal analysis. However, the sampling frequency of most affordable digital cameras is limited to 30–60 Hz, while high-speed cameras for higher frequency vibration measurements are extremely costly. This work develops a computational algorithm capable of performing vibration measurement at a uniform sampling frequency lower than what is required by the Shannon-Nyquist sampling theorem for output-only modal analysis. In particular, the spatio-temporal uncoupling property of the modal expansion of structural vibration responses enables a direct modal decoupling of the temporally-aliased vibration measurements by existing output-only modal analysis methods, yielding (full-field) mode shapes estimation directly. Then the signal aliasing properties in modal analysis is exploited to estimate the modal frequencies and damping ratios. Furthermore, the proposed method is validated by laboratory experiments where output-only modal identification is conducted on temporally-aliased acceleration responses and particularly the temporally-aliased video measurements of bench-scale structures, including a three-story building structure and a cantilever beam.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Flow-dependent tagging of 214 Pb decays in the LZ dark matter detector

The LUX-ZEPLIN (LZ) experiment is searching for dark matter interactions in a liquid xenon time projection chamber (LXe-TPC). This article demonstrates how control of the flow state in the LXe-TPC enables the identification of pairs of sequential alpha-decays, which are used to map fluid flow and ion drift in the liquid target. The resulting transport model is used to tag 214 Pb~ beta-decays, a leading background to dark matter signals in LZ. Temporally evolving volume selections, at a cost of 9.0% of exposure, target the decay of each 214 Pb~ atom up to 81 minutes after production, resulting in (63±6 (s⁢t⁢a⁢t) ±7 (s⁢y⁢s) )% identification of 214 Pb~decays to ground state. We also demonstrate how flow-based tagging techniques enable a novel calibration side band that is concurrent with science data. Lastly we report updated estimates of radon-chain charge branching fractions in liquid xenon, finding branching to 218 Po + at 0.49±0.01, 214 Pb + at 0.48±0.12, and 214 Bi + at 0.74±0.05, with a mean charged ion lifetime in the LZ TPC of 49±4 min.

Lorenzon, Wolfgang [University of Michigan, Ann Ar↗

Evolving Trust in Robots: Specification Through Sequential and Comparative Meta-Analyses

Objective The objectives of this meta-analysis are to explore the presently available empirical findings on the antecedents of trust in robots and use this information to expand upon a previous meta-analytic review of the area. Background Human–robot interaction (HRI) represents an increasingly important dimension of our everyday existence. Currently, the most important element of these interactions is proposed to be whether the human trusts the robot or not. We have identified three overarching categories that exert effects on the expression of trust. These consist of factors associated with (a) the human, (b) the robot, and (c) the context in which any specific HRI event occurs. Method The current body of literature was examined and all qualifying articles pertaining to trust in robots were included in the meta-analysis. A previous meta-analysis on HRI trust was used as the basis for this extended, updated, and evolving analysis. Results Multiple additional factors, which have now been demonstrated to significantly influence trust, were identified. The present results, expressed as points of difference and points of commonality between the current and previous analyses, are identified, explained, and cast in the setting of the emerging wave of HRI. Conclusion The present meta-analysis expands upon previous work and validates the overarching categories of trust antecedent (human-related, robot-related, and contextual), as well as identifying the significant individual precursors to trust within each category. A new and updated model of these complex interactions is offered. Application The identified trust factors can be used in order to promote appropriate levels of trust in robots.

Behavioral Sciences↗