Nordic countries experience post-glacial rebound, a movement where geographical contours slowly change elevation with respect to the mean sea level. The glacial isostatic adjustment (GIA) model aims to explain the phenomena, which combined with seismic data allows geoscientists to reconstruct elastic coefficients and viscosities of the Earth's lithosphere and upper mantle. The use of standard commercial codes are not adequate for GIA simulations and result in significant errors in the displacement field. This negative outcome suggests the development of GIA codes that include advection of pre-stress in the model. The problem set up consists on a solid 2D elastic layer under a flat Earth approximation, described by three different models suggested by current studies in geophysics.For space discretization the mixed finite element method (mFEM) is used and efficient preconditioners are built for the resulting algebraic system in saddle point form. A three level GMRES iterative solution strategy is proposed, based on Schur Complement preconditioners coupled with Multigrid techniques. The implementation is presented as a ready-to-use toolbox that easily deals with problem parameters, geometries, compressible and fully incompressible materials and provides higher accuracy for the displacement field compared with the previously existent codes. It also can be easily extended to 3D geometries and allows the implementation of a viscoelastic mantle. The code is written in C++ using the deal.II library designed for FEM, permitting the use of readily-made software packages, such as Trilinos that are straightforwardly parallelizable.

Due to the continue trend of technology for circuit scaling; optimal sizes for transistors and cables are needed, this scaling comes along with power increase and undesired cross talk effects between the wires which might affect the signal itself. A trade-off curve between power, area and delay is required so that a tuned solution for a specific problem might be found.  Here, an approach based on the dominant time constant optimization technique for modelling of the delay is investigated for an initial square topology with uniform wire and repeater sizing based on convex optimization and Linear Matrix Inequalities. Results with optimal buffer insertion across the bus are also investigated. 

In a Fiber to the Home (FTTH) system, the fiber is connected until the household users. A passive optical network (PON) is a particular type of FTTH networks that uses P2MP (point-to-multipoint fiber premises) in which part of the infrastructure is shared, to allow sending information of multiple channels through one fiber. This study presents a summary of the main concepts about access networks and PON systems and some technologies used to its development, including Time Division Multiplexing (TDM) and Wavelength division multiplexing (WDM), independently or in hybrid configuration (TWDM). From here the key commercial devices for its implementation are identified and modelled, specifying its characteristic parameters to fulfill the minimum requirements of the new international standard ITU-T G.989 for FTTH networks. This project evaluates this standard, which proposes the implementation of NG-PON2 systems at 4x10Gbit/s, that will use the hybrid architecture TWDM and tunable transceivers and ensures coexistence with other legacy PON systems, allocating the downstream and upstream traffic in appropriate wavelength bands without overlapping. The work of analysis is focused in the identification of key devices, such as tunable transmitters and tunable receivers at ONU to achieve the minimum specified requirements. Besides, a first approximation to the coexistence element (CE) is made, for supporting the legacy PON systems. Once the minimum requirements are specified and the key devices identified, the network is designed and built in the commercial simulator VPIphotonicsTM, using CW lasers, external intensity modulators, wavelength multiplexers and the appropriate power splitters. In reception a direct detection scheme is implemented using PIN and APD photodiodes and tunable filters at ONU. Once the design is completed different test are run, the results are analyzed and the design is optimized, in order to dimension the network in terms of fiber reach, number of users, power budget, bandwidth capacity, etc. En un Sistema de Fibra hasta el hogar (FTTH, Fiber to the Home), la fibra es conectada hasta el hogar de los usuarios. Una red óptica pasiva (PON) es un tipo particular de red redes FTTH que utiliza instalaciones punto a multipunto (P2MP), en las cuales se comparte parte de la infraestructura permitiendo enviar información de múltiples canales a través de una sola fibra. Este trabajo presenta un compendio de los principales conceptos sobre redes de acceso y sistemas PON y algunas de las tecnologías usadas para el desarrollo de los mismos, incluyendo multiplexación por división en el tiempo (TDM, Time Division Multiplexing) y multiplexación por división de la longitud de onda (WDM, Wavelength Division Multiplexing), independientes o en configuración híbrida (TWDM). A partir de aquí se identifican y modelan los dispositivos comerciales clave para su implementación, especificando sus parámetros característicos para cumplir con los requerimientos del nuevo estándar internacional ITU-T G.989 para redes FFTH. Este proyecto evalúa dicho estándar, el cual plantea la implementación de sistemas NG-PON2 a 4x10GBit/s, los cuales usarán una arquitectura híbrida TWDM y transceptores sintonizables, además de garantizar la coexistencia con otros sistemas PON heredados, ubicando el tráfico en downstream y upstream en unas bandas de longitud de onda adecuadas sin solapamiento. El trabajo de análisis se enfoca en la identificación de los dispositivos clave, como transmisores y receptores ópticos sintonizables en la ONU para alcanzar los requerimientos mínimos especificados. Además se da una primera aproximación al diseño del elemento de coexistencia (CE, Coexistence element) para el soporte de los sistemas PON heredados. Una vez especificados los requerimientos mínimos e identificados los dispositivos clave, se diseña e implementa la red en el simulador VPIphotonicsTM usando láseres de onda continua, moduladores externos de intensidad, multiplexores de longitud de onda y divisores de potencia apropiados. En recepción se usa detección directa con fotodiodos PIN y APD y filtros ópticos sintonizables en la ONU. Una vez completado el diseño, se realizan diferentes test, se analizan los resultados y se optimiza el diseño para dimensionar la red en términos de número de usuarios, alcance, balance de potencias, capacidad de acho de banda, etc.

The communications services market is undergoing a transition as the growing interconnection of providers is diminishing the importance of coverage as a differentiator, and shifting the focus to the value-added service portfolio and customer experience. At the same time, new options are emerging to deliver optimized technologies for the IT service management framework within the internet sphere.  Orange Business Services, communications service provider, develops and understands together with Accenture consultants, both clients needs and customer experience strategies for deploying an improved management service portal. The process to enable this competitive strategy for OBS includes the research of new tends in technology, the comprehension of customers expectations and needs, and the introduction of the evolutions that competitors are implementing within the market.

In this report we highlighted the significance of the taxicab industry for three main sectors of the economy and, furthermore, to what extent could some changes in labour and taxi regulations be detrimental to the perception of safety by taxicab passengers leading to a reduction in the number of journeys and an economic downturn in several industries. We considered the lack of effectiveness of current applications for giving security to the passengers when they hire a taxi in the street. Taking the taxicab regulation schemes and the issues concerning their passengers as a starting point, we set three groups of security requirements: User safety, Quality service and system’s capabilities and efficiency. The novel TAXI Authentication scheme was created as a response to the need to address those requirements. It is based on the increasingly accepted NFC technology and the tamper-resistant smart cards to construct the Front-office processes, particularly noteworthy is the TAXI Authentication protocol designed as a DDA authentication protocol for efficiency and scalability reasons and the quality service feedback that preserves the anonymity of users facilitating the quality assessment. Whereas, a robust Key Management System supports the Back-office processes grounded in certificates using ECDSA digital signatures due to its efficient performance in smart cards and short-length keys providing a high level of security. We conducted the preliminary security analyst against the proposed security requirements. After that, we applied to real world criminal use-cases concerning the taxicab passengers. The scheme shows promise of combating potential crime under whichever type of taxicab regulation scheme demonstrating great flexibility to adapt to a great number of circumstances and future attacks against the industry.

The scope of this project is the redefinition of the algorithm of the wind-diesel hybrid system for rural electrification using ©IntiGIS, which is a methodology translated into a software developed by CIEMAT, useful for evaluating rural electrification projects that enables to compare different technologies based on the LEC, allowing to find which is the best and less costly technology. The analysis done considers the social and geographical particularities of the study area. The core of the new model is the definition of the renewable fraction for the wind-diesel hybrid system. For this, it was supposed that the renewable fraction will depend in first instance of the wind speed. Here, the objectives are: - find a relationship between the renewable fraction and the wind speed, and if there is a clear relationship, try to express it as a function, and – prove the influence of other parameters such as the fuel price and the consumption load in the renewable fraction behavior. The methodology followed to achieve the point above, was to use ©HOMER to simulate several wind-diesel hybrid systems and size the components of the system, in order to get the optimal renewable fraction in different scenarios. Other important point in the model is the wind capacity factor calculation for medium power wind turbines. As a result of the study, the model is more accurate estimating this factor, due to the fact the study was made, taking into account the power curves of some wind turbines that are in the market and are used for electrification purposes, in order to get an average behavior of the capacity factor when wind speed increases.

Microbial pollution of water is one of the principal causes of life-threatening diarrheal diseases in the developing world. Ultraviolet (UV) light is increasingly recognized as a viable alternative for the disinfection of drinking water and wastewater in developed countries, but its feasibility in low-income areas has to be assessed further. The rural community of Cerro Grande, in Bolivia, has been hit by outbreaks of gastrointestinal diseases, so two UV-based disinfection systems were implemented there. One of them was a fabricated unit, with materials available locally, whereas the other was a commercially-available unit. The fabricated unit was validated following USEPA procedures and was modeled using computational fluid dynamics. It was observed that a UV-based disinfection system can be sustainable for as few as 20 users, and even for 48 users in areas with poor feed water quality and lacking an electrical grid and distribution network, with a monthly cost of US$2.

Abstract: Electric vehicles (EVs) have gained great popularity in recent years in the United States. The reduction in the cost of EVs and the environmental regulations that seek to reduce the polluting gases generated by conventional petrol/diesel cars drive up the demand for electric vehicles. The traditional idea is to consider EVs as energy consumers, however Vehicle-to-Grid (V2G) is a technology that allows the batteries of parked EVs to operate as energy sources. V2G technology allows to store energy in EVs and then release it when this energy is needed. The implementation of V2G technologies can provide benefits to utility companies but also presents a series of challenges that must be analyzed and evaluated. V2G technology has a great potential to help Austin Energy use the energy stored in the EVs. This research focuses on the status of charging systems for EVs in Austin, the basic aspects of V2G technology, the standards related to V2G, and aims to provide specific and valuable applications of V2G technology for Austin Energy.

Resumo Neste trabalho, é apresentado um modelo de programação estocástica (MPE) para resolver o problema de planejamento da expansão da rede de transmissão multiestágio (PERTM). O MPE proposto considera fontes de energia renovável intermitente (FERIs) e a alocação de sistemas de armazenamento de energia (SAEs). A natureza estocástica da demanda e das FERIs é modelada através de períodos representativos (PRs). Os PRs são períodos caraterís-ticos construídos a partir de séries temporais que permitem representar o comportamento do sistema elétrico de potência (SEP). Variações de curto prazo da demanda e das FERIs, a cor-relação horária entre essas variáveis aleatórias, e a sua caracterização de acordo com a regi-ão do SEP a que pertencem são incluídas nos PRs, fornecendo um aceitável esforço compu-tacional para resolver o problema de PERTM. Com base nos PRs, o MPE proposto é formu-lado como um modelo equivalente determinístico de programação linear inteira mista (PLIM) para um conjunto de cenários. O modelo de PLIM incorpora perdas de potência ativa e contingências N-1. Investimento em construção de linhas de transmissão (LTs) e novos SAEs são otimizados. Um método eficiente para reduzir o espaço de busca combinatório e lidar com a dimensionalidade do problema de PERTM foi desenvolvido. Os sistemas de tes-te: Garver de 6 barras e IEEE de 24 barras; e o sistema de transmissão norte-colombiano foram usados para validar o modelo proposto. Os resultados mostram que a otimização si-multânea de LTs e SAEs fornece melhores e mais baratos planos de expansão do que as abordagens tradicionais. A eficácia dos SAEs para deslocar blocos de energia elétrica ao longo dos PRs, adiar investimentos e aliviar o congestionamento da rede é evidenciada. Resumen: En este trabajo se presenta un modelo de programación estocástica (MPE) para resolver el problema de planeamiento multietapa de la expansión de la red de transmisión (PERTM). El MPE propuesto considera las fuentes de energía renovables intermitentes (FERIs) y la instalación de sistemas de almacenamiento de energía (SAEs). La naturaleza estocástica de la demanda y las FERIs se modela a través de períodos representativos (PRs). Los PRs son periodos característicos construidos a partir de series de tiempo que permiten representar el comportamiento del sistema eléctrico de potencia (SEP). Las variaciones de corto plazo en la demanda y las FERIs, la correlación horaria entre estas variables aleatorias y su caracterización según la región del SEP a la que pertenecen se incluyen en los PRs, proporcionando un esfuerzo computacional aceptable para resolver el problema PERTM. Sobre la base de los PRs, el MPE propuesto se formula como un modelo equivalente determinístico de programación lineal entera mixta (PLIM) para un conjunto de escenarios. El modelo PLIM incorpora pérdidas de potencia activa y contingencias N-1. Se optimizan las inversiones en la construcción de líneas de transmisión (LT) y nuevos SAEs. Se desarrolló un método eficiente para reducir el espacio de búsqueda combinatoria y mejorar la eficiencia del modelo. Los sistemas de prueba: Garver de 6 barras y IEEE de 24 barras; y el sistema de transmisión del norte de Colombia fueros usados para validar el modelo propuesto. Los resultados muestran que la optimización simultánea de LTs y SAEs proporciona planes de expansión mejores y más económicos que los enfoques tradicionales. Se evidencia la efectividad de los SAEs en desplazar bloques de energía en el tiempo, posponer inversiones y aliviar la congestión de la red.

Abstract: Infrastructure plays an essential role in the economy and has a positive impact on social stability. In this regard, bridges are a fundamental component of infrastructure; however, most of the bridges currently under operation were constructed a long time ago and have even passed their useful life. Certainly, Flaig & Lark (2000) noted that most of the current bridges in the UK were built between 1,950 and 1,970. Furthermore, Steve Kroft (2014) indicated that almost 70,000 bridges (one out of every nine bridges) in America are structurally deficient. Consequently, it is necessary to develop fast and efficient inspection techniques that meet current requirements, and that could potentially replace the highly- subjective visual observation methods. This research presents a critical comparison among different Machine-Learning algorithms that are currently used in image classification problems and applied them into the automated recognition of cracks in reinforced concrete images so that they can be used as an alternative for the existing inspection methods. Moreover, ensembles with different combinations of the evaluated models were tested as an incentive to improve the performance of individual models. In total, four algorithms that based their operation in mathematical formulations and 10 different architectures of Convolutional Neural Networks were included. The results showed that the Random Forest classifier outperformed the other mathematical classifiers and performed similarly to the best of the neural networks. In addition, it was found that the ensembled models achieved little or no improvement concerning the best of the classifiers included in the stack. Resumen: La infraestructura desempeña un papel esencial en la economía de los países y tiene un efecto positivo en la estabilidad social de las personas. A este respecto, los puentes son un componente fundamental de la infraestructura; sin embargo, la mayoría de los puentes actualmente en funcionamiento se construyeron hace mucho tiempo e incluso han pasado su vida útil. Ciertamente, Flaig y Lark (2000) observaron que la mayoría de los puentes actuales del Reino Unido se construyeron entre 1.950 y 1.970. Además, Steve Kroft (2014) indicó que casi 70.000 puentes (uno de cada nueve puentes) en USA son estructuralmente deficientes. Por consiguiente, es necesario desarrollar técnicas de inspección rápidas y eficientes que cumplan los requisitos actuales y que puedan sustituir a los métodos de inspección visual altamente subjetivos. Esta investigación presenta una comparación crítica entre los diferentes algoritmos de aprendizaje a máquina que se utilizan actualmente en los problemas de clasificación de imágenes y los aplica al reconocimiento automatizado de grietas en las imágenes de concreto reforzado para que puedan utilizarse como alternativa a los métodos de inspección existentes. Además, se ensayaron conjuntos con diferentes combinaciones de los algoritmos de aprendizaje evaluados como incentivo para mejorar el rendimiento de los modelos individuales. En total, se incluyeron cuatro algoritmos que basaban su funcionamiento en formulaciones matemáticas y 10 arquitecturas diferentes de Redes Neuronales Convolucionales. Los resultados mostraron que el clasificador Random Forest superó a los otros clasificadores matemáticos y tuvo un rendimiento similar a la mejor de las redes neuronales. Además, se encontró que los modelos ensamblados lograron poca o ninguna mejora en lo que respecta al mejor de los clasificadores incluidos en el conjunto.