Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “approximation algorithms”

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 577 records · Page 32

A Comprehensive Comparison of Relativistic Particle Integrators

We compare relativistic particle integrators commonly used in plasma physics, showing several test cases relevant for astrophysics. Three explicit particle pushers are considered, namely, the Boris, Vay, and Higuera–Cary schemes. We also present a new relativistic fully implicit particle integrator that is energy conserving. Furthermore, a method based on the relativistic guiding center approximation is included. The algorithms are described such that they can be readily implemented in magnetohydrodynamics codes or Particle-in-Cell codes. Here, our comparison focuses on the strengths and key features of the particle integrators. We test the conservation of invariants of motion and the accuracy of particle drift dynamics in highly relativistic, mildly relativistic, and non-relativistic settings. The methods are compared in idealized test cases, i.e., without considering feedback onto the electrodynamic fields, collisions, pair creation, or radiation. The test cases include uniform electric and magnetic fields, ${\boldsymbol{E}}\times {\boldsymbol{B}}$ fields, force-free fields, and setups relevant for high-energy astrophysics, e.g., a magnetic mirror, a magnetic dipole, and a magnetic null. These tests have direct relevance for particle acceleration in shocks and in magnetic reconnection.

79 ASTRONOMY AND ASTROPHYSICS↗

Lighting energy savings potential of split-pane electrochromic windows controlled for daylighting with visual comfort

A simulation study was conducted to evaluate lighting energy savings of split-pane electrochromic (EC) windows controlled to satisfy key visual comfort parameters. Using the Radiance lighting simulation software, interior illuminance and luminance levels were computed for a south-facing private office illuminated by a window split into two independently-controlled EC panes. The transmittance of these was optimized hourly for a workplane illuminance target while meeting visual comfort constraints, using a least-squares algorithm with linear inequality constraints. Blinds were successively deployed until visual comfort criteria were satisfied. The energy performance of electrochromics proved to be highly dependent on how blinds were controlled. With hourly blind position adjustments, electrochromics showed significantly higher (62percent and 53percent, respectively without and with overhang) lighting energy consumption than clear glass. With a control algorithm designed to better approximate realistic manual control by an occupant, electrochromics achieved significant savings (48percent and 37percent, respectively without and with overhang). In all cases, energy consumption decreased when the workplace illuminance target was increased. In addition, the fraction of time during which the occupant had an unobstructed view of the outside was significantly greater with electrochromics: 10 months out of the year versus a handful of days for the reference case.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Neuromorphic scaling advantages for energy-efficient random walk computations

Neuromorphic computing, which aims to replicate the computational structure and architecture of the brain in synthetic hardware, has typically focused on artificial intelligence applications. What is less explored is whether such brain-inspired hardware can provide value beyond cognitive tasks. Here we show that the high degree of parallelism and configurability of spiking neuromorphic architectures makes them well suited to implement random walks via discrete-time Markov chains. Overall, these random walks are useful in Monte Carlo methods, which represent a fundamental computational tool for solving a wide range of numerical computing tasks. Using IBM’s TrueNorth and Intel’s Loihi neuromorphic computing platforms, we show that our neuromorphic computing algorithm for generating random walk approximations of diffusion offers advantages in energy-efficient computation compared with conventional approaches. We also show that our neuromorphic computing algorithm can be extended to more sophisticated jump-diffusion processes that are useful in a range of applications, including financial economics, particle physics and machine learning.

97 MATHEMATICS AND COMPUTING↗

Fast increased fidelity samplers for approximate Bayesian Gaussian process regression

Gaussian processes (GPs) are common components in Bayesian non-parametric models having a rich methodological literature and strong theoretical grounding. The use of exact GPs in Bayesian models is limited to problems containing several thousand observations due to their prohibitive computational demands. We develop a posterior sampling algorithm using H-matrix approximations that scales at O(n log 2 n). We show that this approximation’s Kullback-Leibler divergence to the true posterior can be made arbitrarily small. Though multidimensional GPs could be used with our algorithm, d-dimensional surfaces are modeled as tensor products of univariate GPs to minimize the cost of matrix construction and maximize computational efficiency. We illustrate the performance of this fast increased fidelity approximate GP, FIFA-GP, using both simulated and non-synthetic data sets

97 MATHEMATICS AND COMPUTING↗

Recent progress in 3-D imaging of sea freight containers

The inspection of very large objects like sea freight containers with X-ray Computed Tomography (CT) is an emerging technology. A complete 3-D CT scan of a see-freight container takes several hours. Of course, this is too slow to apply it to a large number of containers. However, the benefits of a 3-D CT for sealed freight are obvious: detection of potential threats or illicit cargo without being confronted with legal complications or high time consumption and risks for the security personnel during a manual inspection. Recently distinct progress was made in the field of reconstruction of projections with only a relatively low number of angular positions. Instead of today’s 500 to 1000 rotational steps, as needed for conventional CT reconstruction techniques, this new class of algorithms provides the potential to reduce the number of projection angles approximately by a factor of 10. The main drawback of these advanced iterative methods is the high consumption for numerical processing. But as computational power is getting steadily cheaper, there will be practical applications of these complex algorithms in a foreseeable future. In this paper, we discuss the properties of iterative image reconstruction algorithms and show results of their application to CT of extremely large objects scanning a sea-freight container. A specific test specimen is used to quantitatively evaluate the image quality in terms of spatial and contrast resolution and depending on different number of projections.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Controlled Nonlinear Stochastic Delay Equations: Part II: Approximations and Pipe-Flow Representations

This is the second part of a work dealing with key issues that have not been addressed in the modeling and numerical optimization of nonlinear stochastic delay systems. We consider new classes of models, such as those with nonlinear functions of several controls (such as products), each with is own delay, controlled random Poisson measure driving terms, admissions control with delayed retrials, and others. Part I was concerned with issues concerning the class of admissible controls and their approximations, since the classical definitions are inadequate for our models. This part is concerned with transportation equation representations and their approximations. Such representations of nonlinear stochastic delay models have been crucial in the development of numerical algorithms with much reduced memory and computational requirements. The representations for the new models are not obvious and are developed. They also provide a template for the adaptation of the Markov chain approximation numerical methods.

97 MATHEMATICS AND COMPUTING↗

Approximate equations of state in two-temperature plasma mixtures

Approximate thermodynamic state relations for multicomponent atomic and molecular gas mixtures are often constructed by artificially partitioning the mixture into its constituent materials and requiring the separated materials to be in temperature and pressure equilibrium. Iterative numerical algorithms have been employed to enforce this equilibration and compute the resulting approximate state relations in single-temperature mixtures. In partially ionized gas mixtures, there is both theoretical and empirical evidence that equilibrating the chemical potentials, number densities, or partial pressures of the free electrons is likely to produce more accurate results than equilibrating the total pressures. Moreover, in many situations of practical interest the free electrons and heavy particles have different temperatures. In this paper, we present a generalized algorithm for equilibrating the heavy-particle and electron temperatures and a third user-specified independent thermodynamic variable in a two-temperature plasma mixture. Test calculations based on the equilibration of total pressure vs. electron pressure are presented for three different mixtures.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Multi-level methods and approximating distribution functions

Biochemical reaction networks are often modelled using discrete-state, continuous-time Markov chains. System statistics of these Markov chains usually cannot be calculated analytically and therefore estimates must be generated via simulation techniques. There is a well documented class of simulation techniques known as exact stochastic simulation algorithms, an example of which is Gillespie’s direct method. These algorithms often come with high computational costs, therefore approximate stochastic simulation algorithms such as the tau-leap method are used. However, in order to minimise the bias in the estimates generated using them, a relatively small value of tau is needed, rendering the computational costs comparable to Gillespie’s direct method. The multi-level Monte Carlo method (Anderson and Higham, Multiscale Model. Simul. 10:146–179, 2012) provides a reduction in computational costs whilst minimising or even eliminating the bias in the estimates of system statistics. This is achieved by first crudely approximating required statistics with many sample paths of low accuracy. Then correction terms are added until a required level of accuracy is reached. Recent literature has primarily focussed on implementing the multi-level method efficiently to estimate a single system statistic. However, it is clearly also of interest to be able to approximate entire probability distributions of species counts. We present two novel methods that combine known techniques for distribution reconstruction with the multi-level method. We demonstrate the potential of our methods using a number of examples.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

POLARIZED LINE FORMATION WITH LOWER-LEVEL POLARIZATION AND PARTIAL FREQUENCY REDISTRIBUTION

In the well-established theories of polarized line formation with partial frequency redistribution (PRD) for a two-level and two-term atom, it is generally assumed that the lower level of the scattering transition is unpolarized. However, the existence of unexplained spectral features in some lines of the Second Solar Spectrum points toward a need to relax this assumption. There exists a density matrix theory that accounts for the polarization of all the atomic levels, but it is based on the flat-spectrum approximation (corresponding to complete frequency redistribution). In the present paper we propose a numerical algorithm to solve the problem of polarized line formation in magnetized media, which includes both the effects of PRD and the lower level polarization (LLP) for a two-level atom. First we derive a collisionless redistribution matrix that includes the combined effects of the PRD and the LLP. We then solve the relevant transfer equation using a two-stage approach. For illustration purposes, we consider two case studies in the non-magnetic regime, namely, the J {sub a} = 1, J {sub b} = 0 and J {sub a} = J {sub b} = 1, where J {sub a} and J {sub b} represent the total angular momentum quantum numbers of the lower and upper states, respectively. Our studies show that the effects of LLP are significant only in the line core. This leads us to propose a simplified numerical approach to solve the concerned radiative transfer problem.

79 ASTRONOMY AND ASTROPHYSICS↗

A computationally efficient method for tempered fractional differential equations with application

This paper investigates the numerical approximation of the tempered fractional integral by using the Sinc-collocation scheme. The algorithm is extended to solve a class of tempered fractional differential equations that converges to the solution with exponential rate. Several numerical examples compare the numerical approximations with the exact solutions. The behavioral responses of the lumped capacitance model with tempered fractional order for transient conduction are investigated. The efficiency and accuracy of the proposed scheme are analyzed in the perspective of the L{sub 2}-norm error and convergence order.

97 MATHEMATICS AND COMPUTING↗

Counting independent sets using the Bethe approximation

The authors consider the problem of counting the number of independent sets or the partition function of a hard-core model in a graph. The problem in general is computationally hard (P hard). They study the quality of the approximation provided by the Bethe free energy. Belief propagation (BP) is a message-passing algorithm can be used to compute fixed points of the Bethe approximation; however, BP is not always guarantee to converge. As the first result, they propose a simple message-passing algorithm that converges to a BP fixed pont for any grapy. They find that their algorithm converges within a multiplicative error 1 + {var_epsilon} of a fixed point in {Omicron}(n{sup 2}E{sup -4} log{sup 3}(nE{sup -1})) iterations for any bounded degree graph of n nodes. In a nutshell, the algorithm can be thought of as a modification of BP with 'time-varying' message-passing. Next, they analyze the resulting error to the number of independent sets provided by such a fixed point of the Bethe approximation. Using the recently developed loop calculus approach by Vhertkov and Chernyak, they establish that for any bounded graph with large enough girth, the error is {Omicron}(n{sup -{gamma}}) for some {gamma} > 0. As an application, they find that for random 3-regular graph, Bethe approximation of log-partition function (log of the number of independent sets) is within o(1) of corret log-partition - this is quite surprising as previous physics-based predictions were expecting an error of o(n). In sum, their results provide a systematic way to find Bethe fixed points for any graph quickly and allow for estimating error in Bethe approximation using novel combinatorial techniques.

97 MATHEMATICS AND COMPUTING↗

Accuracy versus precision in boosted top tagging with the ATLAS detector

The identification of top quark decays where the top quark has a large momentum transverse to the beam axis, known as top tagging , is a crucial component in many measurements of Standard Model processes and searches for beyond the Standard Model physics at the Large Hadron Collider. Machine learning techniques have improved the performance of top tagging algorithms, but the size of the systematic uncertainties for all proposed algorithms has not been systematically studied. This paper presents the performance of several machine learning based top tagging algorithms on a dataset constructed from simulated proton-proton collision events measured with the ATLAS detector at $\sqrt{s}$ = 13 TeV. The systematic uncertainties associated with these algorithms are estimated through an approximate procedure that is not meant to be used in a physics analysis, but is appropriate for the level of precision required for this study. The most performant algorithms are found to have the largest uncertainties, motivating the development of methods to reduce these uncertainties without compromising performance. To enable such efforts in the wider scientific community, the datasets used in this paper are made publicly available.

47 OTHER INSTRUMENTATION↗

Algorithm 1049: The Delaunay Density Diagnostic

Accurate approximation of a real-valued function depends on two aspects of the available data: the density of inputs within the domain of interest and the variation of the outputs over that domain. There are few methods for assessing whether the density of inputs is sufficient to identify the relevant variations in outputs—i.e., the “geometric scale” of the function—despite the fact that sampling density is closely tied to the success or failure of an approximation method. In this article, we introduce a general purpose, computational approach to detecting the geometric scale of real-valued functions over a fixed domain using a deterministic interpolation technique from computational geometry. The algorithm is intended to work on scalar data in moderate dimensions (2–10). Our algorithm is based on the observation that a sequence of piecewise linear interpolants will converge to a continuous function at a quadratic rate (in L 2 norm) if and only if the data are sampled densely enough to distinguish the feature from noise (assuming sufficiently regular sampling). We present numerical experiments demonstrating how our method can identify feature scale, estimate uncertainty in feature scale, and assess the sampling density for fixed (i.e., static) datasets of input–output pairs. Finally, we include analytical results in support of our numerical findings and have released lightweight code that can be adapted for use in a variety of data science settings.

97 MATHEMATICS AND COMPUTING↗

Quantum simulation of Lindbladian dynamics via repeated interactions

The Lindblad equation generalizes the Schrödinger equation to quantum systems that undergo dissipative dynamics. The quantum simulation of Lindbladian dynamics is therefore non-unitary, preventing a naive application of state-of-the-art quantum algorithms. Here, we make use of an approximate correspondence between Lindbladian dynamics and evolution based on repeated interaction (RI) CPTP maps to write down a Hamiltonian formulation of the Lindblad dynamics and derive a rigorous error bound on the master equation. Specifically, we show that the number of interactions needed to simulate the Liouvillian within error e scales in most physical scenarios as . This is significant because the error in the Lindbladian approximation to the dynamics is not explicitly bounded in existing quantum algorithms for open system simulations. We then provide quantum algorithms to simulate RI maps using an iterative qubitization approach and Trotter–Suzuki formulas, and specifically show that for iterative qubitization the number of operations needed to simulate the dynamics (for a fixed value of ?) scales as in the limit where a0 (the coefficient 1-norm for the system and bath Hamiltonians) asymptotically dominates over the corresponding factor for the interaction Hamiltonian, which is often the case in weak coupling. This scaling would appear to be optimal if the complexity of ? is not considered, which underscores the importance of considering the error in the Liouvillian that we reveal in this work.

Quantum Computing↗

hp -VPINNs: Variational physics-informed neural networks with domain decomposition

We formulate a general framework for hp-variational physics-informed neural networks (hp-VPINNs) based on the nonlinear approximation of shallow and deep neural networks and hp-refinement via domain decomposition and projection onto the space of high-order polynomials. The trial space is the space of neural network, which is defined globally over the entire computational domain, while the test space contains piecewise polynomials. Specifically in this study, the hp-refinement corresponds to a global approximation with a local learning algorithm that can efficiently localize the network parameter optimization. Here, we demonstrate the advantages of hp-VPINNs in both accuracy and training cost for several numerical examples of function approximation and in solving differential equations.

42 ENGINEERING↗

Predicting Intensive Care Unit Length of Stay and Mortality Using Patient Vital Signs: Machine Learning Model Development and Validation

Background: Patient monitoring is vital in all stages of care. In particular, intensive care unit (ICU) patient monitoring has the potential to reduce complications and morbidity, and to increase the quality of care by enabling hospitals to deliver higher-quality, cost-effective patient care, and improve the quality of medical services in the ICU. Objective: We here report the development and validation of ICU length of stay and mortality prediction models. The models will be used in an intelligent ICU patient monitoring module of an Intelligent Remote Patient Monitoring (IRPM) framework that monitors the health status of patients, and generates timely alerts, maneuver guidance, or reports when adverse medical conditions are predicted. Methods: We utilized the publicly available Medical Information Mart for Intensive Care (MIMIC) database to extract ICU stay data for adult patients to build two prediction models: one for mortality prediction and another for ICU length of stay. For the mortality model, we applied six commonly used machine learning (ML) binary classification algorithms for predicting the discharge status (survived or not). For the length of stay model, we applied the same six ML algorithms for binary classification using the median patient population ICU stay of 2.64 days. For the regression-based classification, we used two ML algorithms for predicting the number of days. We built two variations of each prediction model: one using 12 baseline demographic and vital sign features, and the other based on our proposed quantiles approach, in which we use 21 extra features engineered from the baseline vital sign features, including their modified means, standard deviations, and quantile percentages. Results: We could perform predictive modeling with minimal features while maintaining reasonable performance using the quantiles approach. The best accuracy achieved in the mortality model was approximately 89% using the random forest algorithm. The highest accuracy achieved in the length of stay model, based on the population median ICU stay (2.64 days), was approximately 65% using the random forest algorithm. Conclusions: The novelty in our approach is that we built models to predict ICU length of stay and mortality with reasonable accuracy based on a combination of ML and the quantiles approach that utilizes only vital signs available from the patient’s profile without the need to use any external features. This approach is based on feature engineering of the vital signs by including their modified means, standard deviations, and quantile percentages of the original features, which provided a richer dataset to achieve better predictive power in our models.

59 BASIC BIOLOGICAL SCIENCES↗

Smart Grid Integrity Attacks: Characterizations and Countermeasures

Real power injections at loads and generators, and real power flows on selected lines in a transmission network are monitored, transmitted over a SCADA network to the system operator, and used in state estimation algorithms to make dispatch, re-balance and other energy management system [EMS] decisions. Coordinated cyber attacks of power meter readings can be arranged to be undetectable by any bad data detection algorithm. These unobservable attacks present a serious threat to grid operations. Of particular interest are sparse attacks that involve the compromise of a modest number of meter readings. An efficient algorithm to find all unobservable attacks [under standard DC load flow approximations] involving the compromise of exactly two power injection meters and an arbitrary number of power meters on lines is presented. This requires O(n2m) flops for a power system with n buses and m line meters. If all lines are metered, there exist canonical forms that characterize all 3, 4, and 5-sparse unobservable attacks. These can be quickly detected in power systems using standard graph algorithms. Known secure phase measurement units [PMUs] can be used as countermeasures against an arbitrary collection of cyber attacks. Finding the minimum number of necessary PMUs is NP-hard. It is shown that p + 1 PMUs at carefully chosen buses are sufficient to neutralize a collection of p cyber attacks.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Approximate inference on planar graphs using loop calculus and belief progagation

We introduce novel results for approximate inference on planar graphical models using the loop calculus framework. The loop calculus (Chertkov and Chernyak, 2006b) allows to express the exact partition function Z of a graphical model as a finite sum of terms that can be evaluated once the belief propagation (BP) solution is known. In general, full summation over all correction terms is intractable. We develop an algorithm for the approach presented in Chertkov et al. (2008) which represents an efficient truncation scheme on planar graphs and a new representation of the series in terms of Pfaffians of matrices. We analyze in detail both the loop series and the Pfaffian series for models with binary variables and pairwise interactions, and show that the first term of the Pfaffian series can provide very accurate approximations. The algorithm outperforms previous truncation schemes of the loop series and is competitive with other state-of-the-art methods for approximate inference.

97 MATHEMATICS AND COMPUTING↗