Parallel GPU architecture for hyperspectral unmixing based on augmented Lagrangian method

Hyperspectral imaging has become one of the main topics in remote sensing applications, which comprise hundreds of spectral bands at different (almost contiguous) wavelength channels over the same area generating large data volumes comprising several GBs per flight. This high spectral resolution can be used for object detection and for discriminate between different objects based on their spectral characteristics. One of the main problems involved in hyperspectral analysis is the presence of mixed pixels, which arise when the spacial resolution of the sensor is not able to separate spectrally distinct materials. Spectral unmixing is one of the most important task for hyperspectral data exploitation. However, the unmixing algorithms can be computationally very expensive, and even high power consuming, which compromises the use in applications under on-board constraints. In recent years, graphics processing units (GPUs) have evolved into highly parallel and programmable systems. Specifically, several hyperspectral imaging algorithms have shown to be able to benefit from this hardware taking advantage of the extremely high floating-point processing performance, compact size, huge memory bandwidth, and relatively low cost of these units, which make them appealing for onboard data processing. In this paper, we propose a parallel implementation of an augmented Lagragian based method for unsupervised hyperspectral linear unmixing on GPUs using CUDA. The method called simplex identification via split augmented Lagrangian (SISAL) aims to identify the endmembers of a scene, i.e., is able to unmix hyperspectral data sets in which the pure pixel assumption is violated. The efficient implementation of SISAL method presented in this work exploits the GPU architecture at low level, using shared memory and coalesced accesses to memory.

Contact with friction using the augmented Lagrangian Method: a conditional constrained minimization problem

This work presents a formulation of the contact with friction between elastic bodies. This is a non linear problem due to unilateral constraints (inter-penetration of bodies) and friction. The solution of this problem can be found using optimization concepts, modelling the problem as a constrained minimization problem. The Finite Element Method is used to construct approximation spaces. The minimization problem has the total potential energy of the elastic bodies as the objective function, the non-inter-penetration conditions are represented by inequality constraints, and equality constraints are used to deal with the friction. Due to the presence of two friction conditions (stick and slip), specific equality constraints are present or not according to the current condition. Since the Coulomb friction condition depends on the normal and tangential contact stresses related to the constraints of the problem, it is devised a conditional dependent constrained minimization problem. An Augmented Lagrangian Method for constrained minimization is employed to solve this problem. This method, when applied to a contact problem, presents Lagrange Multipliers which have the physical meaning of contact forces. This fact allows to check the friction condition at each iteration. These concepts make possible to devise a computational scheme which lead to good numerical results.

Global minimization using an Augmented Lagrangian method with variable lower-level constraints

A novel global optimization method based on an Augmented Lagrangian framework is introduced for continuous constrained nonlinear optimization problems. At each outer iteration k the method requires the epsilon(k)-global minimization of the Augmented Lagrangian with simple constraints, where epsilon(k) -> epsilon. Global convergence to an epsilon-global minimizer of the original problem is proved. The subproblems are solved using the alpha BB method. Numerical experiments are presented.

Improving ultimate convergence of an augmented Lagrangian method

Optimization methods that employ the classical Powell-Hestenes-Rockafellar augmented Lagrangian are useful tools for solving nonlinear programming problems. Their reputation decreased in the last 10 years due to the comparative success of interior-point Newtonian algorithms, which are asymptotically faster. In this research, a combination of both approaches is evaluated. The idea is to produce a competitive method, being more robust and efficient than its `pure` counterparts for critical problems. Moreover, an additional hybrid algorithm is defined, in which the interior-point method is replaced by the Newtonian resolution of a Karush-Kuhn-Tucker (KKT) system identified by the augmented Lagrangian algorithm. The software used in this work is freely available through the Tango Project web page:http://www.ime.usp.br/similar to egbirgin/tango/.

Non-destructive thermal wave method applied to study thermal properties of fast setting time endodontic cement

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Augmented Lagrangian method with nonmonotone penalty parameters for constrained optimization

At each outer iteration of standard Augmented Lagrangian methods one tries to solve a box-constrained optimization problem with some prescribed tolerance. In the continuous world, using exact arithmetic, this subproblem is always solvable. Therefore, the possibility of finishing the subproblem resolution without satisfying the theoretical stopping conditions is not contemplated in usual convergence theories. However, in practice, one might not be able to solve the subproblem up to the required precision. This may be due to different reasons. One of them is that the presence of an excessively large penalty parameter could impair the performance of the box-constraint optimization solver. In this paper a practical strategy for decreasing the penalty parameter in situations like the one mentioned above is proposed. More generally, the different decisions that may be taken when, in practice, one is not able to solve the Augmented Lagrangian subproblem will be discussed. As a result, an improved Augmented Lagrangian method is presented, which takes into account numerical difficulties in a satisfactory way, preserving suitable convergence theory. Numerical experiments are presented involving all the CUTEr collection test problems.

The boundedness of penalty parameters in an augmented Lagrangian method with constrained subproblems

Augmented Lagrangian methods are effective tools for solving large-scale nonlinear programming problems. At each outer iteration, a minimization subproblem with simple constraints, whose objective function depends on updated Lagrange multipliers and penalty parameters, is approximately solved. When the penalty parameter becomes very large, solving the subproblem becomes difficult; therefore, the effectiveness of this approach is associated with the boundedness of the penalty parameters. In this paper, it is proved that under more natural assumptions than the ones employed until now, penalty parameters are bounded. For proving the new boundedness result, the original algorithm has been slightly modified. Numerical consequences of the modifications are discussed and computational experiments are presented.

Biaxial strain-induced suppression of spinodal decomposition in GaMnAs and GaCrAs

The thermodynamic properties of the magnetic semiconductors GaMnAs and GaCrAs are studied under biaxial strain. The calculations are based on the projector augmented wave method combined with the generalized quasichemical approach to treat the disorder and composition effects. Considering the influence of biaxial strain, we find a tendency to the suppression of binodal decomposition mainly for GaMnAs under compressive strain. For a substrate with a lattice constant 5% smaller than the one of GaAs, for GaMnAs, the solubility limit increases up to 40%. Thus, the strain can be a useful tool for tailoring magnetic semiconductors to the formation or not of embedded nanoclusters. (C) 2010 American Institute of Physics. [doi:10.1063/1.3448025]

Quantum confinement effects on Mn-doped InAs nanocrystals: A first-principles study

We studied the effect of quantum confinement in Mn-doped InAs nanocrystals using theoretical methods. We observe that the stability of the impurities decreases with the size of the nanocrystals, making doping more difficult in small nanoparticles. Substitutional impurities are always more stable than interstitial ones, independent of the size of the nanocrystal. There is also a decrease in the energy difference between the high and low spin configurations, indicating that the critical temperature should decrease with the size of the nanoparticles, in agreement with experimental observations and in detriment to the development of functional spintronic devices with doped nanocrystals. Codoping with acceptors or saturating the nanocrystals with molecules that insert partially empty levels in the energy gap should be an efficient way to increase T(C).

Comparative study of the adsorption and dissociation of vinylacetic acid and acrylic acid on silicon (001)

In this work, we employ the state of the art pseudopotential method, within a generalized gradient approximation to the density functional theory, to investigate the adsorption process of acrylic acid (AAc) and vinylacetic acid (VAA) on the silicon surface. Our total energy calculations support the proposed experimental process, as it indicates that the chemisorption of the molecule is as follows: The gas phase VAA (AAc) adsorbs molecularly to the electrophilic surface Si atom and then dissociates into H(2)C = CH - COO and H, bonded to the electrophilic and nucleophilic surface silicon dimer atoms, respectively. The activation energy for both processes correspond to thermal activations that are smaller than the usual growth temperature. In addition, the electronic structure, calculated vibrational modes, and theoretical scanning tunneling microscopy images are discussed, with a view to contribute to further experimental investigations.

Density functional theory investigation of 3d, 4d, and 5d 13-atom metal clusters

The knowledge of the atomic structure of clusters composed by few atoms is a basic prerequisite to obtain insights into the mechanisms that determine their chemical and physical properties as a function of diameter, shape, surface termination, as well as to understand the mechanism of bulk formation. Due to the wide use of metal systems in our modern life, the accurate determination of the properties of 3d, 4d, and 5d metal clusters poses a huge problem for nanoscience. In this work, we report a density functional theory study of the atomic structure, binding energies, effective coordination numbers, average bond lengths, and magnetic properties of the 3d, 4d, and 5d metal (30 elements) clusters containing 13 atoms, M(13). First, a set of lowest-energy local minimum structures (as supported by vibrational analysis) were obtained by combining high-temperature first- principles molecular-dynamics simulation, structure crossover, and the selection of five well-known M(13) structures. Several new lower energy configurations were identified, e. g., Pd(13), W(13), Pt(13), etc., and previous known structures were confirmed by our calculations. Furthermore, the following trends were identified: (i) compact icosahedral-like forms at the beginning of each metal series, more opened structures such as hexagonal bilayerlike and double simple-cubic layers at the middle of each metal series, and structures with an increasing effective coordination number occur for large d states occupation. (ii) For Au(13), we found that spin-orbit coupling favors the three-dimensional (3D) structures, i.e., a 3D structure is about 0.10 eV lower in energy than the lowest energy known two-dimensional configuration. (iii) The magnetic exchange interactions play an important role for particular systems such as Fe, Cr, and Mn. (iv) The analysis of the binding energy and average bond lengths show a paraboliclike shape as a function of the occupation of the d states and hence, most of the properties can be explained by the chemistry picture of occupation of the bonding and antibonding states.

Density functional theory and ab initio molecular dynamics study of NO adsorption on Pd(111) and Pt(111) surfaces

The origin of the unique geometry for nitric oxide (NO) adsorption on Pd(111) and Pt(111) surfaces as well as the effect of temperature were studied by density functional theory calculations and ab initio molecular dynamics at finite temperature. We found that at low coverage, the adsorption geometry is determined by electronic interactions, depending sensitively on the adsorption sites and coverages, and the effect of temperature on geometries is significant. At coverage of 0.25 monolayer (ML), adsorbed NO at hollow sites prefer an upright configuration, while NO adsorbed at top sites prefer a tilting configuration. With increase in the coverage up to 0.50 ML, the enhanced steric repulsion lead to the tilting of hollow NO. We found that the tilting was enhanced by the thermal effects. At coverage of 0.75 ML with p(2 x 2)-3NO(fcc+hcp+top) structure, we found that there was no preferential orientation for tilted top NO. The interplay of the orbital hybridization, thermal effects, steric repulsion, and their effects on the adsorption geometries were highlighted at the end.

Theory of nitride oxide adsorption on transition metal (111) surfaces: a first-principles investigation

In this work, we report a density functional theory study of nitric oxide (NO) adsorption on close-packed transition metal (TM) Rh(111), Ir(111), Pd(111) and Pt(111) surfaces in terms of adsorption sites, binding mechanism and charge transfer at a coverage of Theta(NO) = 0.25, 0.50, 0.75 monolayer (ML). Based on our study, an unified picture for the interaction between NO and TM(111) and site preference is established, and valuable insights are obtained. At low coverage (0.25 ML), we find that the interaction of NO/TM(111) is determined by an electron donation and back-donation process via the interplay between NO 5 sigma/2 pi* and TM d-bands. The extent of the donation and back-donation depends critically on the coordination number (adsorption sites) and TM d-band filling, and plays an essential role for NO adsorption on TM surfaces. DFT calculations shows that for TMs with high d-band filling such as Pd and Pt, hollow-site NO is energetically the most favorable, and top-site NO prefers to tilt away from the normal direction. While for TMs with low d-band filling (Rh and Ir), top-site NO perpendicular to the surfaces is energetically most favorable. Electronic structure analysis show that irrespective of the TM and adsorption site, there is a net charge transfer from the substrate to the adsorbate due to overwhelming back-donation from the TM substrate to the adsorbed NO molecules. The adsorption-induced change of the work function with respect to bare surfaces and dipole moment is however site dependent, and the work function increases for hollow-site NO, but decreases for top-site NO, because of differences in the charge redistribution. The interplay between the energetics, lateral interaction and charge transfer, which is element dependent, rationalizes the structural evolution of NO adsorption on TM(111) surfaces in the submonolayer regime.

Theoretical investigation of atomic and electronic structures of Ga(2)O(3)(ZnO)(6)

Transparent conducting oxides (TCO) are widely used in technological applications ranging from photovoltaics to thin-film transparent field-effect transistors. In this work we report a first-principles investigation, based on density-functional theory, of the atomic and electronic properties of Ga(2)O(3)(ZnO)(6) (GZO(6)), which is a promising candidate to be used as host oxide for wide band gap TCO applications. We identify a low-energy configuration for the coherent distribution of the Ga and Zn atoms in the cation positions within the experimentally reported orthorhombic GZO(6) structure. Four Ga atoms are located in four-fold sites, while the remaining 12 Ga atoms in the unit cell form four shared Ga agglomerates (a motif of four atoms). The Zn atoms are distributed in the remaining cation sites with effective coordination numbers from 3.90 to 4.50. Furthermore, we identify the natural formation of twin-boundaries in GZO(6), which can explain the zigzag modulations observed experimentally by high-resolution transmission electron microscopy in GZO(n) (n=9). Due to the intrinsic twin-boundary formation, polarity inversion in the ZnO tetrahedrons is present which is facilitated by the formation of the Ga agglomerates. Our analysis shows that the formation of fourfold Ga sites and Ga agglomerates are stabilized by the electronic octet rule, while the distribution of Ga atoms and the formation of the twin-boundary help alleviate excess strain. Finally we identify that the electronic properties of GZO(6) are essentially determined by the electronic properties of ZnO, i.e., there are slight changes in the band gap and optical absorption properties.

Atomistic origins of the phase transition mechanism in Ge(2)Sb(2)Te(5)

The fast and reversible phase transition mechanism between crystalline and amorphous phases of Ge(2)Sb(2)Te(5) has been in debate for several years. Through employing first-principles density functional theory calculations, we identify a direct structural link between the metastable crystalline and amorphous phases. The phase transition is driven by the displacement of Ge atoms along the rocksalt [111] direction from stable octahedron to high energy unstable tetrahedron sites close to the intrinsic vacancy regions, which generates a high energy intermediate phase between metastable and amorphous phases. Due to the instability of Ge at the tetrahedron sites, the Ge atoms naturally shift away from those sites, giving rise to the formation of local-ordered fourfold motifs and the long-range structural disorder. Intrinsic vacancies, which originate from Sb(2)Te(3), lower the energy barrier for Ge displacements, and hence, their distribution plays an important role in the phase transition. The high energy intermediate configuration can be obtained experimentally by applying an intense laser beam, which overcomes the thermodynamic barrier from the octahedron to tetrahedron sites. The high figure of merit of Ge(2)Sb(2)Te(5) is achieved from the optimal combination of intrinsic vacancies provided by Sb(2)Te(3) and the instability of the tetrahedron sites provided by GeTe.