For this paper, we consider the problem of classifying a test sample given incomplete information. This problem arises naturally when data about a test sample is collected over time, or when costs must be incurred to compute the classification features. For example, in a distributed sensor network only a fraction of the sensors may have reported measurements at a certain time, and additional time, power, and bandwidth is needed to collect the complete data to classify. A practical goal is to assign a class label as soon as enough data is available to make a good decision. We formalize this goal through the notion of reliability—the probability that a label assigned given incomplete data would be the same as the label assigned given the complete data, and we propose a method to classify incomplete data only if some reliability threshold is met. Our approach models the complete data as a random variable whose distribution is dependent on the current incomplete data and the (complete) training data. The method differs from standard imputation strategies in that our focus is on determining the reliability of the classification decision, rather than just the class label. We show that the method provides useful reliability estimates of the correctness of the imputed class labels on a set of experiments on time-series data sets, where the goal is to classify the time-series as early as possible while still guaranteeing that the reliability threshold is met.
Graphene, a planar, atomically thin form of carbon, has unique electrical and material properties that could enable new high performance semiconductor devices. Graphene could be of specific interest in the development of room-temperature, high-resolution semiconductor radiation spectrometers. Incorporating graphene into a field-effect transistor architecture could provide an extremely high sensitivity readout mechanism for sensing charge carriers in a semiconductor detector, thus enabling the fabrication of a sensitive radiation sensor. In addition, the field effect transistor architecture allows us to sense only a single charge carrier type, such as electrons. This is an advantage for room-temperature semiconductor radiation detectors, which often suffer from significant hole trapping. Here we report on initial efforts towards device fabrication and proof-of-concept testing. This work investigates the use of graphene transferred onto silicon and silicon carbide, and the response of these fabricated graphene field effect transistor devices to stimuli such as light and alpha radiation.
Sustainability is a critical national security issue for the U.S. and other nations. Sandia National Laboratories (SNL) is already a global leader in sustainability science and technology (SS&T) as documented in this report. This report documents the ongoing work conducted this year as part of the Sustainability Innovation Foundry (SIF). The efforts of the SIF support Sandia's national and international security missions related to sustainability and resilience revolving around energy use, water use, and materials, both on site at Sandia and externally. The SIF leverages existing Sandia research and development (R&D) in sustainability science and technology to support new solutions to complex problems. The SIF also builds on existing Sandia initiatives to support transformation of Sandia into a fully sustainable entity in terms of materials, energy, and water use. In the long term, the SIF will demonstrate the efficacy of sustainability technology developed at Sandia through prototyping and test bed approaches and will provide a common platform for support of solutions to the complex problems surrounding sustainability. Highlights from this year include the Sustainability Idea Challenge, improvements in facilities energy use, lectures and presentations from relevant experts in sustainability [Dr. Barry Hughes, University of Denver], and significant development of the Institutional Transformation (IX) modeling tools to support evaluation of proposed modifications to the SNL infrastructure to realize energy savings.
Material response to dynamic loading is often dominated by microstructure (grain structure, porosity, inclusions, defects). An example critically important to Sandia's mission is dynamic strength of polycrystalline metals where heterogeneities lead to localization of deformation and loss of shear strength. Microstructural effects are of broad importance to the scientific community and several institutions within DoD and DOE; however, current models rely on inaccurate assumptions about mechanisms at the sub-continuum or mesoscale. Consequently, there is a critical need for accurate and robust methods for modeling heterogeneous material response at this lower length scale. This report summarizes work performed as part of an LDRD effort (FY11 to FY13; project number 151364) to meet these needs.
We review the edge element formulation for describing the kinematics of hyperelastic solids. This approach is used to frame the problem of remapping the inverse deformation gradient for Arbitrary Lagrangian-Eulerian (ALE) simulations of solid dynamics. For hyperelastic materials, the stress state is completely determined by the deformation gradient, so remapping this quantity effectively updates the stress state of the material. A method, inspired by the constrained transport remap in electromagnetics, is reviewed, according to which the zero-curl constraint on the inverse deformation gradient is implicitly satisfied. Open issues related to the accuracy of this approach are identified. An optimization-based approach is implemented to enforce positivity of the determinant of the deformation gradient. The efficacy of this approach is illustrated with numerical examples.
Advances in technology for electrochemical energy storage require increased understanding of electrolyte/electrode interfaces, including the electric double layer structure, and processes involved in charging of the interface, and the incorporation of this understanding into quantitative models. Simplified models such as Helmholtz's electric double-layer (EDL) concept don't account for the molecular nature of ion distributions, solvents, and electrode surfaces and therefore cannot be used in predictive, high-fidelity simulations for device design. This report presents theoretical results from models that explicitly include the molecular nature of the electrical double layer and predict critical electrochemical quantities such as interfacial capacitance. It also describes development of experimental tools for probing molecular properties of electrochemical interfaces through optical spectroscopy. These optical experimental methods are designed to test our new theoretical models that provide descriptions of the electric double layer in unprecedented detail.
We report an uncertainty and sensitivity analysis for modeling DC energy from photovoltaic systems. We consider two systems, each comprised of a single module using either crystalline silicon or CdTe cells, and located either at Albuquerque, NM, or Golden, CO. Output from a PV system is predicted by a sequence of models. Uncertainty in the output of each model is quantified by empirical distributions of each model's residuals. We sample these distributions to propagate uncertainty through the sequence of models to obtain an empirical distribution for each PV system's output. We considered models that: (1) translate measured global horizontal, direct and global diffuse irradiance to plane-of-array irradiance; (2) estimate effective irradiance from plane-of-array irradiance; (3) predict cell temperature; and (4) estimate DC voltage, current and power. We found that the uncertainty in PV system output to be relatively small, on the order of 1% for daily energy. Four alternative models were considered for the POA irradiance modeling step; we did not find the choice of one of these models to be of great significance. However, we observed that the POA irradiance model introduced a bias of upwards of 5% of daily energy which translates directly to a systematic difference in predicted energy. Sensitivity analyses relate uncertainty in the PV system output to uncertainty arising from each model. We found that the residuals arising from the POA irradiance and the effective irradiance models to be the dominant contributors to residuals for daily energy, for either technology or location considered. This analysis indicates that efforts to reduce the uncertainty in PV system output should focus on improvements to the POA and effective irradiance models.
Special nuclear material (SNM) detection has applications in nuclear material control, treaty verification, and national security. The neutron and gamma-ray radiation signature of SNMs can be indirectly observed in scintillator materials, which fluoresce when exposed to this radiation. A photomultiplier tube (PMT) coupled to the scintillator material is often used to convert this weak fluorescence to an electrical output signal. The fluorescence produced by a neutron interaction event differs from that of a gamma-ray interaction event, leading to a slightly different pulse in the PMT output signal. The ability to distinguish between these pulse types, i.e., pulse shape discrimination (PSD), has enabled applications such as neutron spectroscopy, neutron scatter cameras, and dual-mode neutron/gamma-ray imagers. In this research, we explore the use of compressive sensing to guide the development of novel mixed-signal hardware for PMT output signal acquisition. Effectively, we explore smart digitizers that extract sufficient information for PSD while requiring a considerably lower sample rate than conventional digitizers. Given that we determine the feasibility of realizing these designs in custom low-power analog integrated circuits, this research enables the incorporation of SNM detection into wireless sensor networks.
We present the Quantum Computer Aided Design (QCAD) simulator that targets modeling quantum devices, particularly silicon double quantum dots (DQDs) developed for quantum qubits. The simulator has three differentiating features: (i) its core contains nonlinear Poisson, effective mass Schrodinger, and Configuration Interaction solvers that have massively parallel capability for high simulation throughput, and can be run individually or combined self-consistently for 1D/2D/3D quantum devices; (ii) the core solvers show superior convergence even at near-zero-Kelvin temperatures, which is critical for modeling quantum computing devices; (iii) it couples with an optimization engine Dakota that enables optimization of gate voltages in DQDs for multiple desired targets. The Poisson solver includes MaxwellBoltzmann and Fermi-Dirac statistics, supports Dirichlet, Neumann, interface charge, and Robin boundary conditions, and includes the effect of dopant incomplete ionization. The solver has shown robust nonlinear convergence even in the milli-Kelvin temperature range, and has been extensively used to quickly obtain the semiclassical electrostatic potential in DQD devices. The self-consistent Schrodinger-Poisson solver has achieved robust and monotonic convergence behavior for 1D/2D/3D quantum devices at very low temperatures by using a predictor-correct iteration scheme. The QCAD simulator enables the calculation of dot-to-gate capacitances, and comparison with experiment and between solvers. It is observed that computed capacitances are in the right ballpark when compared to experiment, and quantum confinement increases capacitance when the number of electrons is fixed in a quantum dot. In addition, the coupling of QCAD with Dakota allows to rapidly identify which device layouts are more likely leading to few-electron quantum dots. Very efficient QCAD simulations on a large number of fabricated and proposed Si DQDs have made it possible to provide fast feedback for design comparison and optimization.
