Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model fitting”

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 1,387 records · Page 77
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Optimization of Energy Storage System Economics and Controls by Incorporating Battery Degradation Costs in REopt

The use of stationary electrochemical energy storage systems utilizing lithium-ion batteries has increased rapidly as the production scale and price for lithium-ion batteries has decreased. These energy storage systems are crucial for maintaining grid resiliency, especially for grids operating with high penetration of renewable energy generation assets or for with a variety of distributed energy generation and storage systems. One challenging factor for the development of battery energy storage systems is estimating the proper sizing, in terms of both power and energy, that minimizes total costs over the lifetime of the systems; this calculation is difficult in simple cases, where a battery is costed independently, but is extremely challenging when building loads and electrical generation by photovoltaic resources are also considered. REopt is a techoeconomic optimization tool developed by NREL to address these challenges. Previously, battery degradation has been priced by simply assuming a 10-year replacement schedule for battery systems. However, this does not account for varying degradation trends observed across real-world batteries, or allow for batteries to be operated in a degradation-aware manner that optimizes battery dispatch based on operating costs. This work incorporates a battery life model into REopt. This battery life model is simple, so that it may be solvable within the constrains of a mixed-integer linear optimization problem, but is fit to accelerated aging data recorded in the lab. To achieve the best possible accuracy for lifetime estimates given these constraints, parameters for the battery life model in REopt are estimated by fitting 20-year simulations of battery life after identifying state-space battery degradation model from accelerated aging data. Comparisons of battery life predicted in REopt and from the state-space battery degradation model to ensure validity of lifetime estimates made by REopt. Battery life and cost is optimized by controlling three decision to minimize system life cost: battery sizing, daily state-of-charge, and daily energy-throughput. The cost of battery degradation as a function of these control variables is then estimated assuming two possible maintenance strategies: replacement, where the entire battery system is replaced if cell reach an end-of-life capacity threshold; and augmentation, which establishes a fund to pay for continual purchase of new batteries to maintain the initial energy capacity of the system. These two strategies offer conservative (for replacement) and optimistic (for augmentation) bounds for total system cost. The degradation cost incurred by these strategies is then used to control battery dispatch decisions, operating the battery in a degradation-aware manner that maximizes battery lifetime while also providing energy when favorable. Because the mixed-integer linear program has perfect foresight of future energy needs, batteries with degradation costs are always operated using 'just-in-time' charging, which is unrealistic, as no energy is left in the storage system to perform other energy services or to serve as emergency back-up power. To combat this, an inequality constraint on the average annual state-of-charge is imposed, and the sensitivity of system cost to average stored energy, e.g., the cost of system resiliency, can be quantified. Analysis of results has several conclusions, for instance, oversizing of battery storage systems is not a cost burden when battery storage is an optimal solution, as any additional battery capacity can simply be utilized to avoid costs of purchasing energy from a utility.

battery↗
DOE OSTI2023

Intra-Grain Local Luminescence Properties of CdSe0.1Te0.9 Thin Films

We report on the local photoluminescence properties of grains and grain boundaries of CdSe 0.1 Te 0.9 thin film deposited by the colossal grain growth method using a wide-field hyperspectral imaging technique. We observed significant variations in the photoluminescence intensity of both individual grains and also that of grain boundaries. Multiple sub-bandgap defect peaks were captured in the luminescence spectra in the energy range 1.2 eV to 1.6 eV. The intensity and peak positions of these sub-gap emissions were slightly different among various grains and at the grain boundaries, revealing intra- and inter-grain variations in these polycrystalline thin films. A recently-developed density-of-states based photoluminescence model was extended to include multiple peaks and was fitted to the data. We observed that at a fixed temperature, the quasi-Fermi level splitting energy and a disordered energy parameter can be extracted locally by use of this model.

data models↗
DOE OSTI2021

Application of machine learning to estimate fireball characteristics and their uncertainty from infrared spectral data

Experiments or events involving high explosives (HE) can be monitored remotely by infrared (IR) sensors to gather information about the configuration or materials involved in the device. Researchers at the Air Force Institute of Technology (AFIT) developed a phenomenological model for HE fireball spectra in the IR range that allows for parameters to be extracted from Fourier transform infrared (FTIR) data. This model includes parameters tied to physical characteristics of the fireball: temperature, size, soot, and gas concentrations. Previous works have sought to recover these parameters by the fitting of either whole spectra or select wavenumber bands to this phenomenological model. Difficulties arise due to the complex relationships between the parameters to be fit. Uncertainty quantification of the estimated fireball parameters is also problematic since HE experiments do not have any ground truth information on the parameters. It is suggested that artificial neural network (ANN) based approaches may be well suited to this problem, because of their ability to capture complex and highly nonlinear relationships. As such, this work seeks to explore the efficacy of deep artificial neural networks (DNNs) for this problem of parameter recovery from spectra and to also investigate the uncertainty of recovering the fireball parameters from FTIR data. Networks are designed using the hyperparameter optimization tool Hyperopt and trained/tested on artificial data generated using the phenomenological model developed by AFIT. The results of applying the network to the artificial data set are compared to a physics-based band approach that uses a selected number of bands based on their physical properties. Information on the uncertainty of estimating parameters from remotely sensed experimental data is obtained by treating the accuracy of the DNN model on artificial data as an upper bound and by examining the impact of emissivity due to soot on parameter estimation error; the results for artificial data are likely to be optimistic as compared to recovering parameters from experimental data.

42 ENGINEERING↗
DOE OSTI2021

Response surface methodology guided adsorption and recovery of free fatty acids from oil using resin

The presence of free fatty acids interferes with the conversion of plant oils to biodiesel. In this work, four strong and weak base resins were evaluated for the removal of free fatty acids (FFA) from oil. Amberlite FPA 51 showed the highest adsorption capacity of FFA. A resin concentration above 3% could enable a higher percentage FFA adsorption. The adsorption process fitted a pseudo-first-order kinetic model and achieved equilibrium in approximately 8 h. A full factorial design was used to optimize the resin and FFA concentrations at a fixed temperature (40°C). A ratio of resin to fatty acid concentrations above 1.875 was sufficient for 70% adsorption and the amount adsorbed continued to increase with further added resin. A two-step washing of resin using hexane and ethanol recovered approximately 67.55%±4.05% of the initially added fatty acid. The resin that was used was regenerated with 5% NaOH and re-used for a minimum of three consecutive cycles. However, the adsorption capacity diminished to 75% of the initial cycle in cycles 2 and 3. Thus, the work presents a resin-based process for deacidification of oil to reduce fatty acid content of oil for biodiesel production.

09 BIOMASS FUELS↗
DOE OSTI2023

How to identify different new neutrino oscillation physics scenarios at DUNE

Next generation neutrino oscillation experiments are expected to measure the remaining oscillation parameters with very good precision. They will have unprecedented capabilities to search for new physics that modify oscillations. DUNE, with its broad band beam, good particle identification, and relatively high energies will provide an excellent environment to search for new physics. If deviations from the standard three-flavor oscillation picture are seen however, it is crucial to know which new physics scenario is found so that it can be verified elsewhere and theoretically understood. We investigate several benchmark new physics scenarios by looking at existing long-baseline accelerator neutrino data from NOvA and T2K and determine at what sensitivity DUNE can differentiate among them. We consider sterile neutrinos and both vector and scalar non-standard neutrino interactions, all with new complex phases, the latter of which could conceivably provide absolute neutrino mass scale information. We find that, in many interesting cases, DUNE will have good model discrimination. We also perform a new fit to NOvA and T2K data with scalar NSI.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Effective climate sensitivity distributions from a 1D model of global ocean and land temperature trends, 1970–2021

Abstract Current theoretically based Earth system models (ESMs) produce Effective Climate Sensitivities (EffCS) that range over a factor of three, with 80% of those models producing stronger global warming trends for 1970–2021 than do observations. To make a more observationally based estimate of EffCS, a 1D time-dependent forcing-feedback model of temperature departures from energy equilibrium is used to match measured ranges of global-average surface and sub-surface land and ocean temperature trends during 1970–2021. In response to two different radiative forcing scenarios, a full range of three model free parameters are evaluated to produce fits to a range of observed surface temperature trends (± 2σ) from four different land datasets and three ocean datasets, as well as deep-ocean temperature trends and borehole-based trend retrievals over land. Land-derived EffCS are larger than over the ocean, and EffCS is lower using the newer Shared Socioeconomic Pathways (SSP245, 1.86 °C global EffCS, ± 34% range 1.48–2.15 °C) than the older Representative Concentration Pathway forcing (RCP6, 2.49 °C global average EffCS, ± 34% range 2.04–2.87 °C). The strongest dependence of the EffCS results is on the assumed radiative forcing dataset, underscoring the role of radiative forcing uncertainty in determining the sensitivity of the climate system to increasing greenhouse gas concentrations from observations alone. The results are consistent with previous observation-based studies that concluded EffCS during the observational period is on the low end of the range produced by current ESMs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

A new database website for nuclear level densities

We introduce a new open-access, web-based database (http://nld.ascsn.net), Current Archive of Nuclear Density of Levels (CANDL), that hosts experimental nuclear level density (NLD) datasets from a variety of techniques and energy ranges. Built using the Dash framework in Python, the database is designed to be interactive and user-friendly, allowing researchers to search, visualize, fit, and export NLD data with minimal effort. This resource includes data extracted from evaporation spectra, Oslo method variants, and other experimental techniques that cover excitation energies beyond the neutron resonance region. The database supports on-the-fly fitting with two widely-used phenomenological models—the Constant Temperature (CT) model and the Back-Shifted Fermi Gas (BSFG) model—selected for their simplicity and computational efficiency. Future versions aim to include additional datasets and model types, as well as easy-to-use interfaces to data science techniques. Here, this platform offers a vital tool for the nuclear physics, astrophysics, medicine, and reactor design communities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

Intrinsic activation energies for ring contraction of allylic cations in zeolites

Cyclic carbocations are important intermediates in zeolite-catalyzed chemistries such as methanol-to-hydrocarbons conversion, naphthenes ring opening, or coke formation. While information about their thermodynamic stability exists, little is known about the kinetics of formation and transformation of cyclic carbocations in zeolites. To fill this knowledge gap, ring contraction of the 1,3,5,5-tetramethylcyclohexenyl cation (C 10 H 17 + ), a representative of 6-membered ring allylic cations, was investigated by in situ UV–vis and IR spectroscopy. Protonic forms of zeolites served as catalysts, at temperatures from 80 °C to 135 °C. Significant oligomerization and hydride transfer in BEA and FAU hampered kinetics analysis, whereas ring contraction dominated in the channels of MOR. The reactant cation, characterized by an electronic absorption at 314 nm and an allylic stretch at 1549 cm −1 , contracted to both a 1,3-alkyl-substituted cyclopentenyl cation (287 nm and 1506 cm −1 ) and a 1,2,3-alkyl-substituted cyclopentenyl cation (297 nm and 1489 cm −1 ). Collection of time-resolved IR spectra and fitting of the intensities with various kinetic models revealed a third transformation, which is expected from thermodynamics: the 1,3-alkyl-substituted cyclopentenyl cation isomerizes to the 1,2,3-alkyl-substituted cyclopentenyl cation. Series of IR spectra recorded at different temperatures delivered intrinsic activation enthalpies (entropies) in MOR of 67 ± 2 kJ mol −1 (−130 ± 6 J mol −1 K −1 ) and 90 ± 3 kJ mol −1 (−70 ± 8 J mol −1 K −1 ) for the contraction to 1,3- and 1,2,3-substituted species, and of 88 ± 5 kJ mol −1 (−84 ± 12 J mol −1 K −1 ) for the isomerization. The findings characterize one path – via contraction of larger rings – to different cyclopentenyl species in zeolites; and the associated, moderate activation energies suggest such transformations contribute to many complex hydrocarbon reaction networks.

Acid catalysis↗
DOE OSTI2024

Exploration of mass splitting and muon/tau mixing parameters for an eV-scale sterile neutrino with IceCube

We present the first three-parameter fit to a 3+1 sterile neutrino model using 7.634 years of data from the IceCube Neutrino Observatory on $ν$ μ + $\overline{ν}$ μ charged-current interactions in the energy range 500–9976 GeV. Our analysis is sensitive to the mass-squared splitting between the heaviest and lightest mass state (Δm$^{2}_{41}$), the mixing matrix element connecting muon flavor to the fourth mass state (|U μ4 | 2 ), and the element connecting tau flavor to the fourth mass state (|U $\tau$4 | 2 ). Predicted propagation effects in matter enhance the signature through a resonance as atmospheric neutrinos from the Northern Hemisphere traverse the Earth to the IceCube detector at the South Pole. The remaining sterile neutrino matrix elements are left fixed, with |U e4 | 2 = 0 and δ 14 = 0, as they have a negligible effect, and δ 24 = π is set to give the most conservative limits. The result is consistent with the no-sterile neutrino hypothesis with a probability of 4.3%. Profiling the likelihood of each parameter yields the 90% confidence levels: 2.4eV 2 < Δm$^{2}_{41}$) < 9.6eV 2 , 0.0081< |U μ4 | 2 < 0.10, and |U $\tau$4 | 2 < 0.035, which narrows the allowed parameter-space for |U $\tau$4 | 2 . However, the primary result of this analysis is the first map of the 3+1 parameter space exploring the interdependence of Δm$^{2}_{41}$, |U μ4 | 2 , and |U $\tau$4 | 2 .

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Optical, structural and aging properties of Al/Sc-based multilayers for the extreme ultraviolet

This manuscript presents the structural characterization of Al/Sc-based periodic multilayer coatings for the extreme ultraviolet (EUV) spectral range. Based on transmission electron microscopy and electron diffraction as well as grazing-incidence and large-angle x-ray diffraction, a model for the layer structure and the interfacial effects of Al/Sc coatings is built. The onset of crystallization in nanoscale Al and Sc layers as a function of thickness is also revealed in these characterizations. The Al/Sc layer model is validated and further refined by fitting in-band and out-of-band EUV reflectance measurements across 5 orders of magnitude in an extended wavelength range from 17 to 80 nm. Here, the same type of EUV reflectance measurements is used to test the Al/Sc aging properties and to demonstrate the spectral response of optimized two- and tri-material multilayers including Al/Sc, Al/Sc/SiC and Mo/Al/Sc.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Curvature Energetics Determined by Alchemical Simulation on Four Topologically Distinct Lipid Phases

The relative curvature energetics of two lipids are tested using thermodynamic integration (TI) on four topologically distinct lipid phases. Simulations use TI to switch between choline headgroup lipids (POPC; that prefers to be flat) and ethanolamine headgroup lipids (POPE; that prefer, for example, the inner monolayer of vesicles). Here, the thermodynamical moving of the lipids between planar, inverse hexagonal (H II ), cubic (Q II ; Pn3m space group), and vesicle topologies reveals differences in material parameters that were previously challenging to access. The methodology allows for predictions of two important lipid material properties: the difference in POPC/POPE monolayer intrinsic curvature (ΔJ 0 ) and the difference in POPC/POPE monolayer Gaussian curvature modulus (Δ$\bar{κ_m}$), both of which are connected to the energetics of topological variation. Analysis of the TI data indicates that, consistent with previous experiment and simulation, the J 0 of POPE is more negative than POPC (ΔJ 0 = –0.018 ± 0.001 Å –1 ). The theoretical framework extracts significant differences in $\bar{κ_m}$ of which POPE is less negative than POPC by 2.0 to 4.0 kcal/mol. The range of these values is determined by considering subsets of the simulations, and disagreement between these subsets suggests separate mechanical parameters at very high curvature. Finally, the fit of the TI data to the model indicates that the position of the pivotal plane of curvature is not constant across topologies at high curvature. Overall, the results offer insights into lipid material properties, the limits of a single HC model, and how to test them using simulation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Probing the Atomic-Scale Structure of Amorphous Aluminum Oxide Grown by Atomic Layer Deposition

Atomic layer deposition (ALD) is a well-established technique for depositing nanoscale coatings with pristine control of film thickness and composition. The trimethylaluminum (TMA) and water (H 2 O) ALD chemistry is inarguably the most widely used and yet to date, we have little information about the atomic-scale structure of the amorphous aluminum oxide (AlO x ) formed by this chemistry. This lack of understanding hinders our ability to establish process-structure-property relationships and ultimately limits technological advancements employing AlOx made via ALD. In this work, we employ synchrotron high-energy X-ray diffraction (HE-XRD) coupled with pair distribution function (PDF) analysis to characterize the atomic structure of amorphous AlO x ALD coatings. We combine ex situ and in operando HE-XRD measurements on ALD AlO x and fit these experimental data using stochastic structural modeling to reveal variations in the Al-O bond length, Al and O coordination environment, and extent of Al vacancies as a function of growth conditions. In particular, the local atomic structure of ALD AlO x is found to change with the substrate and number of ALD cycles. The observed trends are consistent with the formation of bulk Al 2 O 3 surrounded by an O-rich surface layer. We deconvolute these data to reveal atomic-scale structural information for both the bulk and surface phases. Overall, this work demonstrates the usefulness of HE-XRD and PDF analysis in improving our understanding of the structure of amorphous ALD thin films and provides a pathway to evaluate how process changes impact the structure and properties of ALD films.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Néel domain walls with bistable chirality in a perpendicularly magnetized ferrimagnetic insulator

Field-free spin-orbit torque-driven domain wall motion in magnetic thin films with perpendicular magnetic anisotropy (PMA) requires the domain walls to have Néel character. Conventionally, Néel domain walls are stabilized by the Dzyaloshinskii-Moriya interaction (DMI) in ultrathin films. Here, in a europium iron garnet thin film with PMA and an additional uniaxial in-plane anisotropy, we demonstrate two bistable Néel domain wall states in the absence of DMI, and the capability to toggle the wall states with an in-plane field pulse and consequently their directions of motion under a current pulse. We present a phase diagram for the bistable Néel domain wall states as a function of in-plane field pulse width and amplitude. By fitting the experimental data to an analytical model of Néel wall reversal through the nucleation and propagation of Bloch lines, we extract the length of the initial reversed domain wall segment and Bloch line nucleation energy barrier. Current-driven motion of in-plane anisotropy stabilized Néel walls is qualitatively different from that of DMI-stabilized ones owing to the different symmetry of the effective fields that stabilize the Néel configuration. Furthermore, we present a proof of principle demonstration for 2-bit random number generation based on the stochastic reversal of domain wall chirality. These results provide critical insight into the topological energy barrier of Bloch lines and identify paths towards domain wall-based memory and computing devices.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Catalyst and reactor design considerations for selective production of acids by oxidative cleavage of alkenes and unsaturated fatty acids with H 2 O 2

Oxidative cleavage of alkenes and unsaturated fatty acids with hydrogen peroxide gives an efficient and sustainable process to obtain mono- and di-acids for polymers and lubricants with fewer safety risks and less environmental impact than processes that utilize ozone or other inorganic oxidizers (e.g., permanganate, dichromate, etc.). Guided by insight into the mechanisms for competing reaction pathways (i.e., epoxidation of alkene on W–(η 2 -O 2 ) complexes vs. H 2 O 2 decomposition) and the apparent kinetics derived from kinetic experiments, here, we postulate that W-based heterogeneous catalysts can provide high performance and stable operations at low H 2 O 2 concentrations. Semi-batch reactors with continuous introduction of H 2 O 2 solutions offer the means to maintain low H 2 O 2 concentrations while providing sufficient quantities of H 2 O 2 to satisfy the reaction stoichiometry. We derived simple kinetic model equations for the epoxidation, ring-opening, oxidative cleavage, and oxidation steps and fit theses equations to batch experimental data to obtain kinetic parameters. Furthermore, this kinetic model describes the concentration profiles of reactant, oxidant, and products well as shown by agreement with experimental data. Further predictions of the optimal H 2 O 2 feed rate for semi-batch operation utilized by the proposed rate expressions and the reactor design equations suggest that low H 2 O 2 feed rate increases selectivity towards oxidative cleavage products and selective use of H 2 O 2 for oxidative cleavage pathway. Comparisons of oxidative cleavage of 4-octene in batch and semi-batch reactors show that semibatch reactors with optimized molar feed rates of H 2 O 2 increased oxidative cleavage product selectivities (76% to 99%; with an increase in butyric acid selectivity from 1% to 55%) and H 2 O 2 selectivity (3% to 30%). In addition, semibatch reaction conditions used avoid H 2 O 2 -mediated dissolution of W-atoms from the catalyst. Analysis of these findings suggest that solid oxide catalysts will be effective for continuous oxidative cleavage reactions if deployed within fixed-bed reactors that allow for distributed introduction of reactants and therefore low in situ concentrations of H 2 O 2 .

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Optical readout of singlet fission biexcitons in a heteroacene with photoluminescence detected magnetic resonance

Molecular spin systems based on photoexcited triplet pairs formed via singlet fission (SF) are attractive as carriers of quantum information because of their potentially pure and controllable spin polarization, but developing systems that offer optical routes to readout as well as initialization is challenging. Herein, we characterize the electron spin magnetic resonance change in the photoluminescence intensity for a tailored organic molecular crystal while sweeping a microwave drive up to 10 GHz in a broadband loop structure. We observe resonant transitions for both triplet and quintet spin sublevel populations showing their optical sensitivity and revealing the zero-field parameters for each. We map the evolution of these spectra in both microwave frequency and magnetic field, producing a pattern of optically detected magnetic resonance (ODMR) peaks. Fits to these data using a suitable model suggest significant spin polarization in this system with orientation selectivity. Unusual excitation intensity dependence is also observed, which inverts the sign of the ODMR signal for the triplet features, but not for the quintet. These observations demonstrate optical detection of the spin sublevel population dictated by SF and intermolecular geometry, and highlight anisotropic and multi-scale dynamics of triplet pairs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Charge state transition levels of Ni in $β$-Ga 2 O 3 crystals from experiment and theory: An attractive candidate for compensation doping

Nickel-doped β-Ga 2 O 3 crystals were investigated by optical absorption and photoconductivity, revealing Ni-related deep levels. Here, the photoconductivity spectra were fitted using the phenomenological Kopylov and Pikhtin model to identify the energy of the zero-phonon transition (thermal ionization), Franck–Condon shift, and effective phonon energy. The resulting values are compared with the predicted ones by first-principle calculations based on the density functional theory (DFT). An acceptor level (0/-) of 1.9 eV and a donor level (+/0) of 1.1 eV above the valence band minimum are consistently determined for Ni Ga , which preferentially incorporates on the octahedrally coordinated Ga site. Temperature-dependent resistivity measurements yield a thermal activation energy of ~2.0 eV that agrees well with the determined Ni acceptor level. Conclusively, Ni is an eminently suitable candidate for compensation doping for producing semi-insulating β-Ga 2 O 3 substrates due to the position of the acceptor level (below and close to the mid-bandgap).

36 MATERIALS SCIENCE↗
DOE OSTI2024

Tensor decompositions for count data that leverage stochastic and deterministic optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the maximum likelihood estimator (MLE) of the Poisson CPD model. Here, this work presents two new algorithms that extend state-of-the-art local methods for Poisson CPD. Hybrid GCP-CPAPR combines Generalized Canonical Decomposition (GCP) with stochastic optimization and CP Alternating Poisson Regression (CPAPR), a deterministic algorithm, to increase the probability of converging to the MLE over either method used alone. Restarted CPAPR with SVDrop uses a heuristic based on the singular values of the CPD model unfoldings to identify convergence toward optimizers that are not the MLE and restarts within the feasible domain of the optimization problem, thus reducing overall computational cost when using a multi-start strategy. We provide empirical evidence that indicates our approaches outperform existing methods with respect to converging to the Poisson CPD MLE.

CPAPR↗