Search NASA⌕ Search

SEARCH · Search NASA

Results for “algorithm timings”

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 1,495 records · Page 83

Flow simulations about steady-complex and unsteady moving configurations using structured-overlapped and unstructured grids

The limiting factor in simulating flows past realistic configurations of interest has been the discretization of the physical domain on which the governing equations of fluid flow may be solved. In an attempt to circumvent this problem, many Computational Fluid Dynamic (CFD) methodologies that are based on different grid generation and domain decomposition techniques have been developed. However, due to the costs involved and expertise required, very few comparative studies between these methods have been performed. In the present work, the two CFD methodologies which show the most promise for treating complex three-dimensional configurations as well as unsteady moving boundary problems are evaluated. These are namely the structured-overlapped and the unstructured grid schemes. Both methods use a cell centered, finite volume, upwind approach. The structured-overlapped algorithm uses an approximately factored, alternating direction implicit scheme to perform the time integration, whereas, the unstructured algorithm uses an explicit Runge-Kutta method. To examine the accuracy, efficiency, and limitations of each scheme, they are applied to the same steady complex multicomponent configurations and unsteady moving boundary problems. The steady complex cases consist of computing the subsonic flow about a two-dimensional high-lift multielement airfoil and the transonic flow about a three-dimensional wing/pylon/finned store assembly. The unsteady moving boundary problems are a forced pitching oscillation of an airfoil in a transonic freestream and a two-dimensional, subsonic airfoil/store separation sequence. Accuracy was accessed through the comparison of computed and experimentally measured pressure coefficient data on several of the wing/pylon/finned store assembly's components and at numerous angles-of-attack for the pitching airfoil. From this study, it was found that both the structured-overlapped and the unstructured grid schemes yielded flow solutions of comparable accuracy for these simulations. This study also indicated that, overall, the structured-overlapped scheme was slightly more CPU efficient than the unstructured approach.

Newman, James C., III↗

Sequential Principal Component Analysis -An Optimal and Hardware-Implementable Transform for Image Compression

This paper presents the JPL-developed Sequential Principal Component Analysis (SPCA) algorithm for feature extraction / image compression, based on "dominant-term selection" unsupervised learning technique that requires an order-of-magnitude lesser computation and has simpler architecture compared to the state of the art gradient-descent techniques. This algorithm is inherently amenable to a compact, low power and high speed VLSI hardware embodiment. The paper compares the lossless image compression performance of the JPL's SPCA algorithm with the state of the art JPEG2000, widely used due to its simplified hardware implementability. JPEG2000 is not an optimal data compression technique because of its fixed transform characteristics, regardless of its data structure. On the other hand, conventional Principal Component Analysis based transform (PCA-transform) is a data-dependent-structure transform. However, it is not easy to implement the PCA in compact VLSI hardware, due to its highly computational and architectural complexity. In contrast, the JPL's "dominant-term selection" SPCA algorithm allows, for the first time, a compact, low-power hardware implementation of the powerful PCA algorithm. This paper presents a direct comparison of the JPL's SPCA versus JPEG2000, incorporating the Huffman and arithmetic coding for completeness of the data compression operation. The simulation results show that JPL's SPCA algorithm is superior as an optimal data-dependent-transform over the state of the art JPEG2000. When implemented in hardware, this technique is projected to be ideally suited to future NASA missions for autonomous on-board image data processing to improve the bandwidth of communication.

Duong, Tuan A.↗

Genetic Algorithm for Optimization of Neural Networks for Bayesian Inference of Model Uncertainty

The objective of this work was to develop a genetic optimization algorithm that can design a neural network capable of producing uncertainty estimates along with predictions. This algorithm is necessary because the inclusion of uncertainty modeling in a neural network greatly complicates the network’s design space, making the development of a converging model extremely difficult and time consuming. The genetic algorithm presented in this work uses a number of value ranges for various configurable neural network parameters to create a randomly generated population of network architectures. The initially generated population is then evolved over the course of several generations, with the best performing models breeding to produce novel network configurations. Mutations are randomly applied to the network designs to facilitate the development of adaptations beneficial to the task being performed. An experiment was conducted to validate the proposed algorithm, in which the genetic optimizer was tasked with producing a neural network capable of predicting the sound pressure level (SPL) resulting from jet-surface interaction (JSI) noise. The data used for this task was generated at the NASA Glenn Research Center in the Aero-Acoustic Propulsion Laboratory. Starting with an initial population size of 35 randomly generated networks, and evolved over the course of 10 generations, the genetic algorithm produced a design able to predict SPL as a result of JSI noise within 0.272 dB, on average.

Genetic algorithm↗

The Suomi National Polar-Orbiting Partnership (SNPP): Continuing NASA Research and Applications

The Suomi National Polar-orbiting Partnership (SNPP) satellite was successfully launched into a polar orbit on October 28, 2011 carrying 5 remote sensing instruments designed to provide data to improve weather forecasts and to increase understanding of long-term climate change. SNPP provides operational continuity of satellite-based observations for NOAA's Polar-orbiting Operational Environmental Satellites (POES) and continues the long-term record of climate quality observations established by NASA's Earth Observing System (EOS) satellites. In the 2003 to 2011 pre-launch timeframe, NASA's SNPP Science Team assessed the adequacy of the operational Raw Data Records (RDRs), Sensor Data Records (SDRs), and Environmental Data Records (EDRs) from the SNPP instruments for use in NASA Earth Science research, examined the operational algorithms used to produce those data records, and proposed a path forward for the production of climate quality products from SNPP. In order to perform these tasks, a distributed data system, the NASA Science Data Segment (SDS), ingested RDRs, SDRs, and EDRs from the NOAA Archive and Distribution and Interface Data Processing Segments, ADS and IDPS, respectively. The SDS also obtained operational algorithms for evaluation purposes from the NOAA Government Resource for Algorithm Verification, Independent Testing and Evaluation (GRAVITE). Within the NASA SDS, five Product Evaluation and Test Elements (PEATEs) received, ingested, and stored data and performed NASA's data processing, evaluation, and analysis activities. The distributed nature of this data distribution system was established by physically housing each PEATE within one of five Climate Analysis Research Systems (CARS) located at either at a NASA or a university institution. The CARS were organized around 5 key EDRs directly in support of the following NASA Earth Science focus areas: atmospheric sounding, ocean, land, ozone, and atmospheric composition products. The PEATES provided the system level interface with members of the NASA SNPP Science Team and other science investigators within each CARS. A sixth Earth Radiation Budget CARS was established at NASA Langley Research Center (NASA LaRC) to support instrument performance, data evaluation, and analysis for the SNPP Clouds and the Earth's Radiant Budget Energy System (CERES) instrument. Following the 2011 launch of SNPP, spacecraft commissioning, and instrument activation, the NASA SNPP Science Team evaluated the operational RDRs, SDRs, and EDRs produced by the NOAA ADS and IDPS. A key part in that evaluation was the NASA Science Team's independent processing of operational RDRs and SDRs to EDRs using the latest NASA science algorithms. The NASA science evaluation was completed in the December 2012 to April 2014 timeframe with the release of a series of NASA Science Team Discipline Reports. In summary, these reports indicated that the RDRs produced by the SNPP instruments were of sufficiently high quality to be used to create data products suitable for NASA Earth System science and applications. However, the quality of the SDRs and EDRs were found to vary greatly when considering suitability for NASA science. The need for improvements in operational algorithms, adoption of different algorithmic approaches, greater monitoring of on-orbit instrument calibration, greater attention to data product validation, and data reprocessing were prominent findings in the reports. In response to these findings, NASA, in late 2013, directed the NASA SNPP Science Team to use SNPP instrument data to develop data products of sufficiently high quality to enable the continuation of EOS time series data records and to develop innovative, practical applications of SNPP data. This direction necessitated a transition of the SDS data system from its pre-launch assessment mode to one of full data processing and production. To do this, the PEATES, which served as NASA's data product testing environment during the prelaunch and early on-orbit periods, were transitioned to Science Investigator-led Processing Systems (SIPS). The distributed data architecture was maintained in this new system by locating the SIPS at the same institutions at which the CARS and PEATES were located. The SIPS acquire raw SNPP instrument Level 0 (i.e. RDR) data over the full SNPP mission from the NOAA ADS and IDPS through the NASA SDS Data Distribution and Depository Element (SD3E). The SIPS process those data into NASA Level 1, Level 2, and global, gridded Level 3 standard products using peer-reviewed algorithms provided by members of the NASA Science Team. The SIPS work with the NASA SNPP Science Team in obtaining enhanced, refined, or alternate real-time algorithms to support the capabilities of the Direct Readout Laboratory (DRL). All data products, algorithm source codes, coefficients, and auxiliary data used in product generation are archived in an assigned NASA Distributed Active Archive Center (DAAC).

remote sensing↗

Blazing Signature Filter: a library for fast pairwise similarity comparisons

Identifying similarities between datasets is a fundamental task in data mining and has become an integral part of modern scientific investigation. Whether the task is to identify co-expressed genes in large-scale expression surveys or to predict combinations of gene knockouts which would elicit a similar phenotype, the underlying computational task is often a multi-dimensional similarity test. As datasets continue to grow, improvements to the efficiency, sensitivity or specificity of such computation will have broad impacts as it allows scientists to more completely explore the wealth of scientific data. A significant practical drawback of large-scale data mining is the vast majority of pairwise comparisons are unlikely to be relevant, meaning that they do not share a signature of interest. It is therefore essential to efficiently identify these unproductive comparisons as rapidly as possible and exclude them from more time-intensive similarity calculations. The Blazing Signature Filter (BSF) is a highly efficient pairwise similarity algorithm which enables extensive data mining within a reasonable amount of time. The algorithm transforms datasets into binary metrics, allowing it to utilize the computationally efficient bit operators and provide a coarse measure of similarity. As a result, the BSF can scale to high dimensionality and rapidly filter unproductive pairwise comparison. Furthermore, two bioinformatics applications of the tool are presented to demonstrate the ability to scale to billions of pairwise comparisons and the usefulness of this approach.

59 BASIC BIOLOGICAL SCIENCES↗

PSLQ: An Algorithm to Discover Integer Relations

Let x = (x{sub 1}, x{sub 2} {hor_ellipsis}, x{sub n}) be a vector of real or complex numbers. x is said to possess an integer relation if there exist integers a{sub i}, not all zero, such that a{sub 1}x{sub 1} + a{sub 2}x{sub 2} + {hor_ellipsis} + a{sub n}x{sub n} = 0. By an integer relation algorithm, we mean a practical computational scheme that can recover the vector of integers ai, if it exists, or can produce bounds within which no integer relation exists. As we will see in the examples below, an integer relation algorithm can be used to recognize a computed constant in terms of a formula involving known constants, or to discover an underlying relation between quantities that can be computed to high precision. At the present time, the most effective algorithm for integer relation detection is the 'PSLQ' algorithm of mathematician-sculptor Helaman Ferguson [10, 4]. Some efficient 'multi-level' implementations of PSLQ, as well as a variant of PSLQ that is well-suited for highly parallel computer systems, are given in [4]. PSLQ constructs a sequence of integer-valued matrices B{sub n} that reduces the vector y = xB{sub n}, until either the relation is found (as one of the columns of B{sub n}), or else precision is exhausted. At the same time, PSLQ generates a steadily growing bound on the size of any possible relation. When a relation is found, the size of smallest entry of the vector y abruptly drops to roughly 'epsilon' (i.e. 10{sup -p}, where p is the number of digits of precision). The size of this drop can be viewed as a 'confidence level' that the relation is real and not merely a numerical artifact - a drop of 20 or more orders of magnitude almost always indicates a real relation. Very high precision arithmetic must be used in PSLQ. If one wishes to recover a relation of length n, with coefficients of maximum size d digits, then the input vector x must be specified to at least nd digits, and one must employ nd-digit floating-point arithmetic. Maple and Mathematica include multiple precision arithmetic facilities and Maple ships with a full implementation of PSLQ. One may also use any of several freeware multiprecision software packages, for example the ARPREC package by the first author and colleagues at LBNL [7]. In the remaining sections we describe various representative applications of PSLQ. More detail about these examples is given in [8] and the references therein.

97 MATHEMATICS AND COMPUTING↗

High Gain Antenna Pointing on the Mars Exploration Rovers

This paper describes the algorithm used to point the high gain antennae on NASA/JPL's Mars exploration rovers. Each rover's gimballed antenna must track the Earth as it moves across the Martian sky during communication sessions. The pointing algorithm accounts for obstacles to the line-of-sight posed by (1) features on the rover and in the surrounding environment (2) gimbal range limitations, and (3) kinematic singularities in the gimbal mechanism. The algorithm treats all obstacles with a generalized approach that computes the intercept-times to each obstacle. Where possible, the algorithm takes advantage of pairs of joint-space solutions arising from the mechanism design. The algorithm chooses the solution that provides the longest obstruction-free tracking time. Upon encountering an obstacle, the algorithm automatically switches to the other solution if it is not also obstructed. This algorithm has successfully provided obstruction-free pointing for both rovers throughout the mission.

inverse kinematics↗

A fast and memory-sparing probabilistic selection algorithm for the GPU

A fast and memory-sparing probabilistic top-N selection algorithm is implemented on the GPU. This probabilistic algorithm gives a deterministic result and always terminates. The use of randomization reduces the amount of data that needs heavy processing, and so reduces both the memory requirements and the average time required for the algorithm. This algorithm is well-suited to more general parallel processors with multiple layers of memory hierarchy. Probabilistic Las Vegas algorithms of this kind are a form of stochastic optimization and can be especially useful for processors having a limited amount of fast memory available.

99 GENERAL AND MISCELLANEOUS↗

Leveraging Randomized Compiling for the QITE Algorithm

The success of the current generation of Noisy Intermediate-Scale Quantum (NISQ) hardware shows that quantum hardware may be able to tackle complex problems even without error correction. One outstanding issue is that of coherent errors arising from the increased complexity of these devices. These errors can accumulate through a circuit, making their impact on algorithms hard to predict and mitigate. Iterative algorithms like Quantum Imaginary Time Evolution are susceptible to these errors. This article presents the combination of both noise tailoring using Randomized Compiling and error mitigation with a purification. We also show that Cycle Benchmarking gives an estimate of the reliability of the purification. We apply this method to the Quantum Imaginary Time Evolution of a Transverse Field Ising Model and report an energy estimation and a ground state infidelity both below 1\%. Our methodology is general and can be used for other algorithms and platforms. We show how combining noise tailoring and error mitigation will push forward the performance of NISQ devices.

Ville, Jean-Loup↗

Iterative and Self-Consistent Estimation of Degradation and Soiling Loss in PV Systems - A Case Study

In many parts of the world, soiling is a significant loss mechanism for PV energy production. Methods for accurately quantifying historical and current soiling losses are needed. Furthermore, state-of-the-art methods to estimate degradation rates based on production data do not handle soiling in a satisfactory manner. In order to accurately estimate either of these quantities, it is in many cases helpful to consider their combined effect on PV system performance. In this work we make use of the Combined Degradation and Soiling (CODS) algorithm, which iteratively decomposes time series of daily normalized energy yield into four components: a soiling ratio, a degradation trend, seasonal variations, and noise. We demonstrate the use of the model on field performance data from a utility-scale power plant in a high-soiling environment in the Middle East. We apply the algorithm to the temperature-corrected performance index of the DC power on inverter level, quantifying how soiling varies over time and through the plant. The irradiation-weighted average soiling loss in this plant is 2.3 %, with a median soiling rate of -0.086 %/day. Between the inverters in the site in study, estimated degradation rates range between -0.86 %/year and -0.15 %/year, with a mean rate of -0.48 %/year.

41 EE - Solar Energy Technologies Office (EE-4S)↗

Design and simulation of the snowflake divertor control for NSTX-U

This study presents the development of a physics-based multiple-input-multiple-output algorithm for real-time feedback control of snowflake divertor (SFD) configurations on the National Spherical Torus eXperiment Upgrade (NSTX-U). A model of the SFD configuration response to applied voltages on the divertor control coils is first derived and then used, in conjunction with multivariable control synthesis techniques, to design an optimal state feedback controller for the configuration. To demonstrate the capabilities of the controller, a nonlinear simulator for axisymmetric shape control was developed for NSTX-U which simultaneously evolves the currents in poloidal field coils based upon a set of feedback-computed voltage commands, calculates the induced currents in passive conducting structures, and updates the plasma equilibrium by solving the free-boundary Grad-Shafranov problem. Closed-loop simulations demonstrate that the algorithm enables controlled operations in a variety of SFD configurations and provides capabilities for accurate tracking of time-dependent target trajectories for the divertor geometry. In particular, simulation results suggest that a time-varying controller which can properly account for the evolving SFD dynamical response is not only desirable but necessary for achieving acceptable control performance. The algorithm presented in this paper has been implemented in the NSTX-U Plasma Control System in preparation for future control and divertor physics experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Design and simulation of the snowflake divertor control for NSTX-U

This paper presents the development of a physics-based multiple-input-multiple-output algorithm for real-time feedback control of snowflake divertor (SFD) configurations on the National Spherical Torus eXperiment Upgrade (NSTX-U). A model of the SFD configuration response to applied voltages on the divertor control coils is first derived and then used, in conjunction with multivariable control synthesis techniques, to design an optimal state feedback controller for the configuration. To demonstrate the capabilities of the controller, a nonlinear simulator for axisymmetric shape control was developed for NSTX-U which simultaneously evolves the currents in poloidal field coils based upon a set of feedback-computed voltage commands, calculates the induced currents in passive conducting structures, and updates the plasma equilibrium by solving the free-boundary Grad-Shafranov problem. Closed-loop simulations demonstrate that the algorithm enables controlled operations in a variety of SFD configurations and provides capabilities for accurate tracking of time-dependent target trajectories for the divertor geometry. In particular, simulation results suggest that a time-varying controller which can properly account for the evolving SFD dynamical response is not only desirable but necessary for achieving acceptable control performance. The algorithm presented in this paper has been implemented in the NSTX-U Plasma Control System in preparation for future control and divertor physics experiments.

snowflake divertor, MIMO feedback control, nonline↗

RESOLVING THE ELECTROCHEMICAL EQUATIONS OF A SOLID OXIDE FUEL CELL FOR USE IN TRANSIENT SIMULATION AND INTEGRATION INTO CYBER-PHYSICAL SYSTEMS

A major challenge with complex cyber-physical systems stems from long model computational time that creates a mismatch between the model system and the physical system. The numerical modeling of solid oxide fuel cells (SOFCs) presents particular challenges due to the highly coupled nature of the underlying equations and the multiphysics needed to fully resolve their behavior during a transient event. To this end current approaches revolve around splitting the computational efforts into resolving temperature effects and resolving electrochemical effects. Current methods employed for the transient simulation of an SOFC for implementation in the Hybrid Performance (HyPer) facility cyberphysical plant at the National Energy Technology Laboratory reveal a distinct need for accelerated results with a high degree of stability. To this aim, an investigation into the computational time for the code reveals that the underlying electrochemical algorithm takes an order of magnitude more time than its thermal counterpart and has a tendency to vary in terms of iteration time and as such a rework of the underlying system is proposed. The primary method for accelerated electrochemical algorithm solutions is to employ higher order root finding recipes for the resolution of the highly coupled electrochemical equations. This is done with the intention to reduce the overall number of subiterations necessary for resolving voltage, current density, and species concentration, properties of the fuel cell that are all directly coupled and require nested iterative approaches. The overall objective of this approach is an order of magnitude reduction in calculation time without sacrificing stability and increasing accuracy. Specific approaches involve using both bounded and unbounded techniques, such as the False Position method and the Secant method (or if applicable Newton-Raphson) respectively, the drawbacks being slower convergence for False Position and instability for the Secant or Newton-Raphson methods. Current preliminary results on simplified versions of the parent functions involved for electrochemical calculations indicate a reduction in computational steps by a factor of two for the secant method and a factor of three for Newton-Raphson. When implemented into new modified electrochemical algorithms, the results indicate a possible order of magnitude reduction in calculation time.

Arias, Jesus↗

Dynamic Power Network State Estimation with Asynchronous Measurements

The operation of distribution networks is becoming increasingly volatile, due to fast variations of renewables and, hence, net-loading conditions. To perform a reliable state estimation under these conditions, this paper considers the case where measurements from meters, phasor measurement units, and distributed energy resources are collected and processed in real time to produce estimates of the state at a fast time scale. Streams of measurements collected in real time and at heterogenous rates render the underlying processing asynchronous, and poses severe strains on workhorse state estimation algorithms. In this work, a real-time state estimation algorithm is proposed, where data are processed on the fly. Starting from a regularized least-squares model, and leveraging appropriate linear models, the proposed scheme boils down to a linear dynamical system where the state is updated based on the previous estimate and on the measurement gathered from a few available sensors. The estimation error is shown to be always bounded under mild condition. Numerical simulations are provided to corroborate the analytical findings.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Dynamic Power Network State Estimation with Asynchronous Measurements: Preprint

The operation of distribution networks is becoming increasingly volatile, due to fast variations of renewables and, hence, net-loading conditions. To perform a reliable state estimation under these conditions, this paper considers the case where measurements from meters, phasor measurement units, and distributed energy resources are collected and processed in real time to produce estimates of the state at a fast time scale. Streams of measurements collected in real time and at heterogenous rates render the underlying processing asynchronous, and poses severe strains on workhorse state estimation algorithms. In this work, a real-time state estimation algorithm is proposed, where data are processed on the fly. Starting from a regularized least-squares model, and leveraging appropriate linear models, the proposed scheme boils down to a linear dynamical system where the state is updated based on the previous estimate and on the measurement gathered from a few available sensors. The estimation error is shown to be always bounded under mild condition. Numerical simulations are provided to corroborate the analytical findings.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Development of Multistep and Degenerate Variational Integrators for Applications in Plasma Physics

Geometric integrators yield high-fidelity numerical results by retaining conservation laws in the time advance. A particularly powerful class of geometric integrators is symplectic integrators, which are widely used in orbital mechanics and accelerator physics. An important application presently lacking symplectic integrators is the guiding center motion of magnetized particles represented by non-canonical coordinates. Because guiding center trajectories are foundational to many simulations of magnetically confined plasmas, geometric guiding center algorithms have high potential for impact. The motivation is compounded by the need to simulate long-pulse fusion devices, including ITER, and opportunities in high performance computing, including the use of petascale resources and beyond. This dissertation uses a systematic procedure for constructing geometric integrators - known as variational integration - to deliver new algorithms for guiding center trajectories and other plasma-relevant dynamical systems. These variational integrators are non-trivial because the Lagrangians of interest are degenerate - the Euler-Lagrange equations are first-order differential equations and the Legendre transform is not invertible. The first contribution of this dissertation is that variational integrators for degenerate Lagrangian systems are typically multistep methods. Multistep methods admit parasitic mode instabilities that can ruin the numerical results. These instabilities motivate the second major contribution: degenerate variational integrators. By replicating the degeneracy of the continuous system, degenerate variational integrators avoid parasitic mode instabilities. The new methods are therefore robust geometric integrators for degenerate Lagrangian systems. These developments in variational integration theory culminate in one-step degenerate variational integrators for non-canonical magnetic field line flow and guiding center dynamics. The guiding center integrator assumes coordinates such that one component of the magnetic field is zero; it is shown how to construct such coordinates for nested magnetic surface configurations. Additionally, collisional drag effects are incorporated in the variational guiding center algorithm for the first time, allowing simulation of energetic particle thermalization. Advantages relative to existing canonical-symplectic and non-geometric algorithms are numerically demonstrated. All algorithms have been implemented as part of a modern, parallel, ODE-solving library, suitable for use in high-performance simulations.

Guiding Center↗

Aeroelastic Flight Data Analysis with the Hilbert-Huang Algorithm

This report investigates the utility of the Hilbert Huang transform for the analysis of aeroelastic flight data. It is well known that the classical Hilbert transform can be used for time-frequency analysis of functions or signals. Unfortunately, the Hilbert transform can only be effectively applied to an extremely small class of signals, namely those that are characterized by a single frequency component at any instant in time. The recently-developed Hilbert Huang algorithm addresses the limitations of the classical Hilbert transform through a process known as empirical mode decomposition. Using this approach, the data is filtered into a series of intrinsic mode functions, each of which admits a well-behaved Hilbert transform. In this manner, the Hilbert Huang algorithm affords time-frequency analysis of a large class of signals. This powerful tool has been applied in the analysis of scientific data, structural system identification, mechanical system fault detection, and even image processing. The purpose of this report is to demonstrate the potential applications of the Hilbert Huang algorithm for the analysis of aeroelastic systems, with improvements such as localized online processing. Applications for correlations between system input and output, and amongst output sensors, are discussed to characterize the time-varying amplitude and frequency correlations present in the various components of multiple data channels. Online stability analyses and modal identification are also presented. Examples are given using aeroelastic test data from the F-18 Active Aeroelastic Wing airplane, an Aerostructures Test Wing, and pitch plunge simulation.

Brenner, Martin J.↗

CP decomposition for tensors via alternating least squares with QR decomposition

The CP tensor decomposition is used in applications such as machine learning and signal processing to discover latent low-rank structure in multidimensional data. Computing a CP decomposition via an alternating least squares (ALS) method reduces the problem to several linear least squares problems. The standard way to solve these linear least squares subproblems is to use the normal equations, which inherit special tensor structure that can be exploited for computational efficiency. However, the normal equations are sensitive to numerical ill-conditioning, which can compromise the results of the decomposition. In this paper, we develop versions of the CP-ALS algorithm using the QR decomposition and the singular value decomposition, which are more numerically stable than the normal equations, to solve the linear least squares problems. Our algorithms utilize the tensor structure of the CP-ALS subproblems efficiently, have the same complexity as the standard CP-ALS algorithm when the input is dense and the rank is small, and are shown via examples to produce more stable results when ill-conditioning is present. Our MATLAB implementation achieves the same running time as the standard algorithm for small ranks, and we show that the new methods can obtain lower approximation error.

97 MATHEMATICS AND COMPUTING↗