No description available.

[ES] La presente tesis doctoral tiene como objetivo fundamental el desarrollo de nuevos componentes pasivos de microondas para bandas altas de frecuencia, que puedan llegar a operar, al menos, hasta la banda Ka. Estos nuevos componentes surgen en respuesta a la clara tendencia en el ámbito de las comunicaciones vía satélite de aumentar la frecuencia y el ancho de banda de trabajo, en aras de incrementar las velocidades de transmisión. Concretamente, se ha abordado el diseño de elementos básicos de un subsistema de comunicaciones vía satélite, como son las uniones en guía de onda y los filtros paso banda. Por una parte, se han desarrollado uniones compensadas en guía de onda double ridge (T en plano H, T en plano E y T mágica), superando las limitaciones en ancho de banda que presentan las uniones tradicionales construidas sobre guía rectangular. Se ha pretendido lograr el objetivo de desarrollar unas uniones capaces de interconectar elementos que cubran un amplio ancho de banda, lo que también incluye los filtros de nueva generación presentados en esta tesis. Por otro lado, se ha propuesto una novedosa técnica para el diseño de filtros paso banda con inversores de inmitancia, basada en el slope parameter. A diferencia de la técnica tradicional, este nuevo proceso tiene en cuenta la dependencia real de los elementos del filtro con la frecuencia, calculando y actualizando el slope parameter de cada resonador cargado con sus elementos de acoplo adyacentes en un proceso iterativo que llega a convergencia en pocas iteraciones. Así, este procedimiento basado en simulaciones electromagnéticas y cálculos, mejora la equivalencia entre el prototipo circuital y la estructura real sintetizada, ya que tiene en cuenta no sólo la dependencia con la frecuencia de los resonadores y los inversores individuales, sino también los efectos de carga que existen entre ellos. Además, se ha identificado que el mapeado en frecuencia elegido ha de ser el correcto para cada estructura y especificaciones del filtro. Como resultado, se obtienen unas mejores dimensiones iniciales, reduciendo de esta manera el coste computacional del paso final de optimización, que puede incluso llegar a ser innecesario en algunos casos. Este nuevo procedimiento se ha ilustrado mediante la síntesis de dos filtros de modo evanescente y mediante filtros radiantes que emplean una topología similar. Finalmente, se ha llevado a cabo el estudio de una nueva familia de filtros denominada Modified Hybrid Folded Rectangular Waveguide (MHFRW), que además de operar en las citadas bandas de frecuencia, son capaces de manejar altos niveles de potencia, son fáciles de diseñar y fabricar (sin requerir tornillos de sintonía en muchos casos) y además presentan bajas pérdidas. Todo ello con la ventaja añadida de poder poner un número de ceros de transmisión mayor al orden del filtro (filtros superelípticos) mediante el uso de stubs cortocircuitados de longitud adecuada. Para ilustrar su funcionamiento se han presentado dos ejemplos detallados. El primero de ellos, que no presenta ceros de transmisión, se ha diseñado mediante la nueva técnica del slope parameter. El segundo se trata de un filtro superelíptico, con 4 ceros de transmisión (2 antes de la banda de paso y 2 después) implementados con 4 stubs en un filtro de orden 3. Por último, se ha mostrado la gran potencialidad práctica de esta familia de filtros mediante su aplicación en complejos multiplexores para bancos de medida de intermodulación pasiva (PIM). / [CAT] La present tesi doctoral té com a objectiu fonamental el desenvolupament de nous components passius de microones per a bandes altes de freqüència, que puguen arribar a operar almenys fins a la banda Ka. Aquests nous components sorgeixen en resposta a la clara tendència en l'àmbit de les comunicacions via satèl·lit d'augmentar la freqüència i l'amplada de banda de treball, a fi de incrementar les velocitats de transmissió. Concretament, s'ha abordat el disseny d'elements bàsics d'un subsistema de comunicacions via satèl·lit, com són les unions en guia d'ona i els filtres passabanda. D'una banda, s'han desenvolupat unions compensades en guia d'ona double ridge (T en pla H, T en pla E i T màgica), superant així les limitacions d'amplada de banda que presenten les unions tradicionals, construïdes sobre guia rectangular. S'ha pretès aconseguir l'objectiu de desenvolupar unes unions capaces d'interconnectar elements que cobrisquen una gran amplada de banda, cosa que inclou els filtres de nova generació presentats en aquesta tesi. D'altra banda, s'ha proposat una nova tècnica per al disseny de filtres passabanda amb inversors d'immitància, basada en l'slope parameter. A diferència de la tècnica tradicional, aquest nou procés té en compte la dependència real dels elements del filtre respecte de la freqüència, i calcula i actualitza l'slope parameter de cada ressonador carregat amb els elements d'acoblament adjacents, en un procés iteratiu que arriba a convergència en poques iteracions. Així, aquest procediment basat en simulacions electromagnètiques i càlculs millora l'equivalència entre el prototip circuital i l'estructura real sintetitzada, ja que té en compte no sols la dependència respecte de la freqüència dels ressonadors i els inversors individuals, sinó també els efectes de càrrega que hi ha entre aquests. A més, s'ha identificat que el mapatge en freqüència triat ha de ser el correcte per a cada estructura i especificacions del filtre. Com a resultat, s'obtenen unes millors dimensions inicials, i d'aquesta manera es redueix el cost computacional del pas final d'optimització, que fins i tot pot arribar a ser innecessari en alguns casos. Aquest nou procediment s'ha il·lustrat mitjançant la síntesi de dos filtres de mode evanescent i mitjançant filtres radiants que empren una topologia similar. Finalment, s'ha portat a terme l'estudi de una nova família de filtres denominada Modified Hybrid Folded Rectangular Waveguide (MHFRW), que, a més d'operar en les citades bandes de freqüència, són capaços de manejar alts nivells de potència, fàcils de dissenyar i fabricar (sense requerir caragols de sintonia en molts casos) i, a més, presenten baixes pèrdues. Tot això, amb l'avantatge afegit de poder posar un nombre de zeros de transmissió major que l'ordre del fltre (filtres superel·líptics) mitjançant l'ús de stubs curtcircuitats de longitud adequada. Per il·lustrar-ne el funcionament se n'han presentat dos exemples detallats. El primer d'aquests, que no presenta zeros de transmissió, s'ha dissenyat mitjançant la nova tècnica de l'slope parameter. El segon es tracta de un filtre superel·líptic, amb 4 zeros de transmissió (2 abans de la banda de pas i 2 després) implementats amb 4 slope en un filtre d'ordre 3. Finalment, s'ha mostrat la gran potencialitat pràctica d'aquesta família de filtres mitjançant l'aplicació en multiplexors complexos per a bancs de mesurament de la intermodulació passiva (PIM). / [EN] The main objective of this Ph.D. thesis is the development of new passive microwave components for high frequency bands, able to work at least up to Ka band. These new components aim at satisfying the needs of new communication systems, especially for satellite communications, where there is a trend to increase frequency and bandwidth of operation. Specifically, the design of basic hardware elements of satellite communication subsystems has been addressed, such as waveguide junctions and bandpass filters. On the one hand, compensated T-junctions have been developed in double ridge waveguide (H-plane T, E-plane T and magic T), in order to overcome the bandwidth limitations of traditional junctions built on rectangular waveguide technology. The development of junctions able of interconnecting elements over a wide bandwidth has been pretended, including also the new generation of filters presented in this Ph.D. thesis. On the other hand, a novel technique for the design of bandpass filters with immittance inverters has been proposed. This technique is based on the slope parameter. Unlike traditional techniques, this new process takes into account the actual dependence of filter elements with frequency, calculating and updating the slope parameter of each resonator loaded with its adjacent coupling elements in an iterative process. This process reaches convergence after few iterations. Therefore, this procedure (based on electromagnetic simulations and calculations) improves the equivalence between the circuit prototype and the real structure, since it takes into account the frequency dependence of resonators and individual inverters, including also the loading effects between them. In addition, it is important to point out that an appropriate frequency mapping must be chosen for each structure and filter specifications. As a result, better initial dimensions can be obtained, thus reducing the computational cost of the final optimization step, which can even become unnecessary in some cases. This new procedure has been illustrated by the synthesis of two evanescent-mode filters and several radiating filters based on a similar topology. Finally, the study of a new filter family called Modified Hybrid Folded Rectangular Waveguide (MHFRW) has been carried out. Besides operating in the mentioned frequency bands, theses filters are capable of handling high power levels, are easy to design and manufacture (without requiring tuning screws in many cases) and also present low losses. Likewise, this type of filters are able to generate a number of transmission zeros greater than the order of the filter (i.e., provide super-elliptic responses) by means of shortcircuited stubs. To illustrate its operation, two examples have been presented in detail. The first of them, which does not have transmission zeros, has been implemented using the new technique of slope parameter. Accordingly, the suitability of this method for this type of structure is demonstrated. The second is a super-elliptic filter, with 4 transmission zeros (2 before the passband and 2 after it) implemented with 4 stubs in a filter of order 3. To conclude and show the practical capabilities of the proposed family of filters, they have been successfully applied in the development of complex multiplexers for passive intermodulation (PIM) measurement test benches. / Ruiz Garnica, J. (2019). Desarrollo de Nuevos Componentes Pasivos de Microondas en Tecnología Guiada para Bandas Altas de Frecuencia [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/133058 / TESIS

In this thesis, we present selected methods of regression and classification analysis in terms of robust optimization which aim to compensate for data imprecisions and measurement errors. In the first part, ordinary least squares method and its generalizations derived within the context of robust optimization - ridge regression and Lasso method are introduced. The connection between robust least squares and stated generalizations is also shown. Theoretical results are accompanied with simulation study investigating from a different perspective the robustness of stated methods. In the second part, we define a modern classification method - Support Vector Machines (SVM). Using the obtained knowledge, we formulate a robust SVM method, which can be applied in robust classification. The final part is devoted to the biometric identification of a style of typing and an individual based on keystroke dynamics using the formulated theory. Powered by TCPDF (www.tcpdf.org)

Thesis (Ph. D.)--Joint Program in Oceanography (Massachusetts Institute of Technology, Dept. of Earth, Atmospheric and Planetary Sciences, and the Woods Hole Oceanographic Institution), 2001. / Page 250 blank. / Includes bibliographical references. / Half of the ocean crust is formed at spreading centers with total opening rates less than 40 km/Myr. The objective of this Thesis is to investigate temporal variations in active ridge processes and crustal aging at slow-spreading centers by comparing axial crustal structure with that on conjugate flanks of the slow-spreading Mid-Atlantic Ridge (MAR) (full rate, 20 km/Myr) and the ultra-slow spreading Southwest Indian Ridge (SWIR) (full rate, 14 km/Myr). Seismic refraction data collected along the rift valley and flanking rift mountains of the OH-1 segment (35ʻN) at the MAR show that the entire crustal section is constructed within a zone that is less than 5 km wide. Shallow-level hydrothermal circulation within the axial valley is suggested by the rift mountain seismic profiles, which show that the upper crust is 20% thinner and 16% faster along strike than zero-age crust. These effects probably result from fissure sealing within the extrusive crust. Deeper crustal velocities remain relatively constant at the segment midpoint within the first 2 Myr, but are reduced near the segment offsets presumably by faulting and fracturing associated with uplift out of the rift valley. / (cont.) A temporal variation in axial melt supply is suggested by a 15% difference in along-strike crustal thickness between the rift valley and rift mountains, with relatively less melt supplied today than 2 Ma. Crustal accretion at the SWIR appears to occur in a similar manner as at the MAR, although gravity and seismic data indicate that the average crustal thickness is 2-4 km less at theultra-slow spreading SWIR. A 25 Myr record on both flanks of the ridge shows that seafloor spreading has been highly asymmetric through time, with 35% faster crustal accretion on the Antarctic (south) plate. A small-offset non-transform discontinuity between two ridge segments is just as stable as two neighboring transform discontinuities, although a single mantle Bouguer gravity anomaly centered over the non-transform offset indicates that this boundary does not significantly perturb underlying mantle flow. Off-axis magnetic anomalies are recorded with high fidelity despite the very low spreading rates and the absence of a basaltic upper crust in one area. The lower crust may be the dominant off-axis carrier of the magnetic signal, contrary to traditionalmodels of crustal magnetic structure. Morphological and gravity data show evidence of asymmetric crustal accretion across the SWIR ridge axis, with slightly warmer mantletemperatures beneath the slower-spreading African (north) plate. / by Allegra Hosford. / Ph.D.

No description available.

Optical flow is the apparent motion of brightness patterns within an image scene. Algorithms used to calculate the optical flow for a sequence of images are useful in a variety of applications, including motion detection and obstacle avoidance. Typical optical flow algorithms are computationally intense and run slowly when implemented in software, which is problematic since many potential applications of the algorithm require real-time calculation in order to be useful. To increase performance of the calculation, optical flow has recently been implemented on FPGA and GPU platforms. These devices are able to process optical flow in real-time, but are generally less accurate than software solutions. For this thesis, two different optical flow algorithms have been implemented to run on a GPU using NVIDIA's CUDA SDK. Previous FPGA implementations of the algorithms exist and are used to make a comparison between the FPGA and GPU devices for the optical flow calculation. The first algorithm calculates optical flow using 3D gradient tensors and is able to process 640x480 images at about 238 frames per second with an average angular error of 12.1 degrees when run on a GeForce 8800 GTX GPU. The second algorithm uses increased smoothing and a ridge regression calculation to produce a more accurate result. It reduces the average angular error by about 2.3x, but the additional computational complexity of the algorithm also reduces the frame rate by about 1.5x. Overall, the GPU outperforms the FPGA in frame rate and accuracy, but requires much more power and is not as flexible. The most significant advantage of the GPU is the reduced design time and effort needed to implement the algorithms, with the FPGA designs requiring 10x to 12x the effort.

The vertebrate embryonic limb is a key model in elucidating the genetic basis underlying the three dimensional morphogenesis of structures. Despite the wealth of insights that have been generated from this model, many long-standing questions remain. For example, it has been known for over 70 years that the apical ectodermal ridge (AER) of the embryonic limb is essential for distal outgrowth and patterning of the adjacent limb mesenchyme. The mechanisms whereby the AER does accomplish outgrowth and patterning are still poorly understood. We propose that secreted FGFs from the AER activate Wnt5a expression in gradient fashion, which in turn provides an instructional cue to direct outgrowth in the direction of increasing Wnt5a expression (i.e. toward the distal tip of the limb). In vivo and in vitro models were used to test this hypothesis. We placed Wnt5a expressing L-cell implants into stage 23 chick limb buds and demonstrate that labeled mesenchyme cells grow toward the source of Wnt5a. Purified Wnt5a soaked heparin bead implants have only a marginal effect on directed growth of the adjacent mesenchyme, whereas a greater effect was seen with beads soaked in Wnt5a conditioned media. Using an in vitro model where cultured limb mesenchyme cells were subjected to a gradient of conditioned Wnt5a media or purified Wnt5a, we show no specific migratory direction. However, clusters of cells tended to move toward the source of Wnt5a indicating that it might be necessary for the cells to be in complete contact to respond to the Wnt5a signal. Taken together, our results suggest that Wnt5a is sufficient to direct limb mesenchyme. This finding has given support to a new model of limb development proposed by our lab and referred to as the Mesenchyme Recruitment Model.

The study of the instabilities in boiling water reactors is of significant importance to the safety withwhich they can be operated, as they can cause damage to the reactor posing risks to both equipmentand personnel. The instabilities that concern this paper are progressive growths in the oscillatingpower of boiling-water reactors. As thermal power is oscillatory is important to be able to identifywhether or not the power amplitude is stable. The main focus of this paper has been the development of a neural network estimator of these insta-bilities, fitting a non-linear model function to data by estimating it’s parameters. In doing this, theambition was to optimize the networks to the point that it can deliver near ”best-guess” estimationsof the parameters which define these instabilities, evaluating the usefulness of these networks whenapplied to problems like this. The goal was to design both MLP(Multi-Layer Perceptron) and SVR/KRR(Support Vector Regres-sion/Kernel Rigde Regression) networks and improve them to the point that they provide reliableand useful information about the waves in question. This goal was accomplished only in part asthe SVR/KRR networks proved to have some difficulty in ascertaining the phase shift of the waves.Overall, however, these networks prove very useful in this kind of task, succeeding with a reasonabledegree of confidence to calculating the different parameters of the waves studied.

Internal waves are generated in stratified fluids, like the ocean, where density increases with depth. Tides are one of the major generation mechanisms of internal waves. As the tides move water back and forth over underwater topography, internal waves can be generated. Topography slope and amplitude are major factors in the behavior of the generated internal wave field. In order to further understand the effects topographic shape plays, the effect of asymmetry on internal waves is investigated. This research investigates internal waves generated by dual-ridge topographies. Four cases of symmetric topographies, T, M, W, and W2, with three different peak spacings are compared to their singular ridge counter parts at three oscillation frequencies, ω = 0.6N, ω = 0.75N, and ω = 0.9N. Both subcritical and supercritical symmetric ridges were investigated. Experiments were also performed for subcritical, asymmetric dual ridges at the middle oscillation frequency. The internal wave fields were captured with synthetic schlieren and analyzed with the Hilbert transform and sum of kinetic energy in wavenumber space. It is found that for wave fields from substantially separated ridges, mixing and wave interference occurs that decreases total kinetic energy of the system.

No description available.