A dynamic factor model for mid-term forecasting of wind power generation

The main objective of this paper is the development and application of multivariate time series models for forecasting aggregated wind power production in a country or region. Nowadays, in Spain, Denmark or Germany there is an increasing penetration of this kind of renewable energy, somehow to reduce energy dependence on the exterior, but always linked with the increaseand uncertainty affecting the prices of fossil fuels. The disposal of accurate predictions of wind power generation is a crucial task both for the System Operator as well as for all the agents of the Market. However, the vast majority of works rarely onsider forecasting horizons longer than 48 hours, although they are of interest for the system planning and operation. In this paper we use Dynamic Factor Analysis, adapting and modifying it conveniently, to reach our aim: the computation of accurate forecasts for the aggregated wind power production in a country for a forecasting horizon as long as possible, particularly up to 60 days (2 months). We illustrate this methodology and the results obtained for real data in the leading country in wind power production: Denmark

Biosensors based on vertically interrogated optofluidic sensing cells

El objetivo de la presente tesis doctoral es el desarrollo de un nuevo concepto de biosensor óptico sin marcado, basado en una combinación de técnicas de caracterización óptica de interrogación vertical y estructuras sub-micrométricas fabricadas sobre chips de silicio. Las características más importantes de dicho dispositivo son su simplicidad, tanto desde el punto de vista de medida óptica como de introducción de las muestras a medir en el área sensible, aspectos que suelen ser críticos en la mayoría de sensores encontrados en la literatura. Cada uno de los aspectos relacionados con el diseño de un biosensor, que son fundamentalmente cuatro (diseño fotónico, caracterización óptica, fabricación y fluídica/inmovilización química) son desarrollados en detalle en los capítulos correspondientes. En la primera parte de la tesis se hace una introducción al concepto de biosensor, en qué consiste, qué tipos hay y cuáles son los parámetros más comunes usados para cuantificar su comportamiento. Posteriormente se realiza un análisis del estado del arte en la materia, enfocado en particular en el área de biosensores ópticos sin marcado. Se introducen también cuáles son las reacciones bioquímicas a estudiar (inmunoensayos). En la segunda parte se describe en primer lugar cuáles son las técnicas ópticas empleadas en la caracterización: Reflectometría, Elipsometría y Espectrometría; además de los motivos que han llevado a su empleo. Posteriormente se introducen diversos diseños de las denominadas "celdas optofluídicas", que son los dispositivos en los que se va a producir la interacción bioquímica. Se presentan cuatro dispositivos diferentes, y junto con ellos, se proponen diversos métodos de cálculo teórico de la respuesta óptica esperada. Posteriormente se procede al cálculo de la sensibilidad esperada para cada una de las celdas, así como al análisis de los procesos de fabricación de cada una de ellas y su comportamiento fluídico. Una vez analizados todos los aspectos críticos del comportamiento del biosensor, se puede realizar un proceso de optimización de su diseño. Esto se realiza usando un modelo de cálculo simplificado (modelo 1.5-D) que permite la obtención de parámetros como la sensibilidad y el límite de detección de un gran número de dispositivos en un tiempo relativamente reducido. Para este proceso se escogen dos de las celdas optofluídicas propuestas. En la parte final de la tesis se muestran los resultados experimentales obtenidos. En primer lugar, se caracteriza una celda basada en agujeros sub-micrométricos como sensor de índice de refracción, usando para ello diferentes líquidos orgánicos; dichos resultados experimentales presentan una buena correlación con los cálculos teóricos previos, lo que permite validar el modelo conceptual presentado. Finalmente, se realiza un inmunoensayo químico sobre otra de las celdas propuestas (pilares nanométricos de polímero SU-8). Para ello se utiliza el inmunoensayo de albumina de suero bovino (BSA) y su anticuerpo (antiBSA). Se detalla el proceso de obtención de la celda, la funcionalización de la superficie con los bioreceptores (en este caso, BSA) y el proceso de biorreconocimiento. Este proceso permite dar una primera estimación de cuál es el límite de detección esperable para este tipo de sensores en un inmunoensayo estándar. En este caso, se alcanza un valor de 2.3 ng/mL, que es competitivo comparado con otros ensayos similares encontrados en la literatura. La principal conclusión de la tesis es que esta tipología de dispositivos puede ser usada como inmunosensor, y presenta ciertas ventajas respecto a los actualmente existentes. Estas ventajas vienen asociadas, de nuevo, a su simplicidad, tanto a la hora de medir ópticamente, como dentro del proceso de introducción de los bioanalitos en el área sensora (depositando simplemente una gota sobre la micro-nano-estructura). Los cálculos teorícos realizados en los procesos de optimización sugieren a su vez que el comportamiento del sensor, medido en magnitudes como límite de detección biológico puede ser ampliamente mejorado con una mayor compactación de pilares, alcanzandose un valor mínimo de 0.59 ng/mL). The objective of this thesis is to develop a new concept of optical label-free biosensor, based on a combination of vertical interrogation optical techniques and submicron structures fabricated over silicon chips. The most important features of this device are its simplicity, both from the point of view of optical measurement and regarding to the introduction of samples to be measured in the sensing area, which are often critical aspects in the majority of sensors found in the literature. Each of the aspects related to the design of biosensors, which are basically four (photonic design, optical characterization, fabrication and fluid / chemical immobilization) are developed in detail in the relevant chapters. The first part of the thesis consists of an introduction to the concept of biosensor: which elements consists of, existing types and the most common parameters used to quantify its behavior. Subsequently, an analysis of the state of the art in this area is presented, focusing in particular in the area of label free optical biosensors. What are also introduced to study biochemical reactions (immunoassays). The second part describes firstly the optical techniques used in the characterization: reflectometry, ellipsometry and spectrometry; in addition to the reasons that have led to their use. Subsequently several examples of the so-called "optofluidic cells" are introduced, which are the devices where the biochemical interactions take place. Four different devices are presented, and their optical response is calculated by using various methods. Then is exposed the calculation of the expected sensitivity for each of the cells, and the analysis of their fabrication processes and fluidic behavior at the sub-micrometric range. After analyzing all the critical aspects of the biosensor, it can be performed a process of optimization of a particular design. This is done using a simplified calculation model (1.5-D model calculation) that allows obtaining parameters such as sensitivity and the detection limit of a large number of devices in a relatively reduced time. For this process are chosen two different optofluidic cells, from the four previously proposed. The final part of the thesis is the exposition of the obtained experimental results. Firstly, a cell based sub-micrometric holes is characterized as refractive index sensor using different organic fluids, and such experimental results show a good correlation with previous theoretical calculations, allowing to validate the conceptual model presented. Finally, an immunoassay is performed on another typology of cell (SU-8 polymer pillars). This immunoassay uses bovine serum albumin (BSA) and its antibody (antiBSA). The processes for obtaining the cell surface functionalization with the bioreceptors (in this case, BSA) and the biorecognition (antiBSA) are detailed. This immunoassay can give a first estimation of which are the expected limit of detection values for this typology of sensors in a standard immunoassay. In this case, it reaches a value of 2.3 ng/mL, which is competitive with other similar assays found in the literature. The main conclusion of the thesis is that this type of device can be used as immunosensor, and has certain advantages over the existing ones. These advantages are associated again with its simplicity, by the simpler coupling of light and in the process of introduction of bioanalytes into the sensing areas (by depositing a droplet over the micro-nano-structure). Theoretical calculations made in optimizing processes suggest that the sensor Limit of detection can be greatly improved with higher compacting of the lattice of pillars, reaching a minimum value of 0.59 ng/mL).

Valoración de recursos propios en empresas mineras: evidencia y comparación con otros sectores

El capital financiero es muy volátil y si el inversor no obtiene una remuneración adecuada al riesgo que asume puede plantearse el retirar su capital del patrimonio de la empresa y, en consecuencia, producir un cambio estructural en cualquier sector de la economía. El objetivo principal es el estudio de los coeficientes de regresión (coeficiente beta) de los modelos de valoración de activos empleados en Economía Financiera, esto es, el estudio de la variación de la rentabilidad de los activos en función de los cambios que suceden en los mercados. La elección de los modelos utilizados se justifica por la amplia utilización teórica y empírica de los mismos a lo largo de la historia de la Economía Financiera. Se han aplicado el modelo de valoración de activos de mercado (capital asset pricing model, CAPM), el modelo basado en la teoría de precios de arbitraje (arbitrage pricing theory, APT) y el modelo de tres factores de Fama y French (FF). Estos modelos se han aplicado a los rendimientos mensuales de 27 empresas del sector minero que cotizan en la bolsa de Nueva York (New York Stock Exchange, NYSE) o en la de Londres (London Stock Exchange, LSE), con datos del período que comprende desde Enero de 2006 a Diciembre de 2010. Los resultados de series de tiempo y sección cruzada tanto para CAPM, como para APT y FF producen varios errores, lo que sugiere que muchas empresas del sector no han podido obtener el coste de capital. También los resultados muestran que las empresas de mayor riesgo tienden a tener una menor rentabilidad. Estas conclusiones hacen poco probable que se mantenga en el largo plazo el equilibrio actual y puede que sea uno de los principales factores que impulsen un cambio estructural en el sector minero en forma de concentraciones de empresas. ABSTRACT Financial capital is highly volatile and if the investor does not get adequate compensation for the risk faced he may consider withdrawing his capital assets from the company and consequently produce a structural change in any sector of the economy. The main purpose is the study of the regression coefficients (beta) of asset pricing models used in financial economics, that is, the study of variation in profitability of assets in terms of the changes that occur in the markets. The choice of models used is justified by the extensive theoretical and empirical use of them throughout the history of financial economics. Have been used the capital asset pricing model, CAPM, the model XII based on the arbitrage pricing theory (APT) and the three-factor model of Fama and French (FF). These models have been applied to the monthly returns of 27 mining companies listed on the NYSE (New York Stock Exchange) or LSE(London Stock Exchange), using data from the period covered from January 2006 to December 2010. The results of time series and cross sectional regressions for CAPM, APT and FF produce some errors, suggesting that many companies have failed to obtain the cost of capital. Also the results show that higher risk firms tend to have lower profitability. These findings make it unlikely to be mainteined over the long term the current status and could drive structural change in the mining sector in the form of mergers.

A shake table test of typical mediterranean reinforced concrete structures

This research investigates the ultimate earthquake resistance of typical RC moment resisting frames designed accordingly to current standards, in terms of ultimate energy absorption/dissipation capacity. Shake table test of a 2/5 scale model, under several intensities of ground motion, are carried out. The loading effect of the earthquake is expressed as the total energy that the quake inputs to the structure, and the seismic resistance is interpreted as the amount of energy that the structure dissipates in terms of cumulative inelastic strain energy.

Modelling and forecasting fossil fuels, CO2 and electricity prices and their volatilities

In the current uncertain context that affects both the world economy and the energy sector, with the rapid increase in the prices of oil and gas and the very unstable political situation that affects some of the largest raw materials’ producers, there is a need for developing efficient and powerful quantitative tools that allow to model and forecast fossil fuel prices, CO2 emission allowances prices as well as electricity prices. This will improve decision making for all the agents involved in energy issues. Although there are papers focused on modelling fossil fuel prices, CO2 prices and electricity prices, the literature is scarce on attempts to consider all of them together. This paper focuses on both building a multivariate model for the aforementioned prices and comparing its results with those of univariate ones, in terms of prediction accuracy (univariate and multivariate models are compared for a large span of days, all in the first 4 months in 2011) as well as extracting common features in the volatilities of the prices of all these relevant magnitudes. The common features in volatility are extracted by means of a conditionally heteroskedastic dynamic factor model which allows to solve the curse of dimensionality problem that commonly arises when estimating multivariate GARCH models. Additionally, the common volatility factors obtained are useful for improving the forecasting intervals and have a nice economical interpretation. Besides, the results obtained and methodology proposed can be useful as a starting point for risk management or portfolio optimization under uncertainty in the current context of energy markets.

CSM-IXIM: A New Maize Simulation Model for DSSAT Version 4.5

Th e CERES-Maize model is the most widely used maize (Zea mays L.) model and is a recognized reference for comparing new developments in maize growth, development, and yield simulation. Th e objective of this study was to present and evaluate CSMIXIM, a new maize simulation model for DSSAT version 4.5. Code from CSM-CERES-Maize, the modular version of the model, was modifi ed to include a number of model improvements. Model enhancements included the simulation of leaf area, C assimilation and partitioning, ear growth, kernel number, grain yield, and plant N acquisition and distribution. Th e addition of two genetic coeffi cients to simulate per-leaf foliar surface produced 32% smaller root mean square error (RMSE) values estimating leaf area index than did CSM-CERES. Grain yield and total shoot biomass were correctly simulated by both models. Carbon partitioning, however, showed diff erences. Th e CSM-IXIM model simulated leaf mass more accurately, reducing the CSM-CERES error by 44%, but overestimated stem mass, especially aft er stress, resulting in similar average RMSE values as CSM-CERES. Excessive N uptake aft er fertilization events as simulated by CSM-CERES was also corrected, reducing the error by 16%. Th e accuracy of N distribution to stems was improved by 68%. Th ese improvements in CSM-IXIM provided a stable basis for more precise simulation of maize canopy growth and yield and a framework for continuing future model developments

Variation of PM2.5 concentrations in relation to street washing activities

Several studies conducted in urban areas have pointed out that road dust resuspension contributes significantly to PM concentration levels. Street washing is one of the methods proposed to reduce resuspended road dust contributions to ambient PM concentrations. As resuspended particles are mainly found in the coarse mode, published studies investigating the effects of street washing have focused on PM10 size fraction. As the PM2.5 mass fraction of particles originating from mechanical abrasion processes may still be significant we conducted a study in order to evaluate the effects of street washing on the mitigation of resuspension of fine particles. The PM2.5 mass concentration data were examined and integrated with the occurrence of street washing activities. In addition, the effect of the meteorological variability, traffic flow and street washing activities, on ambient PM2.5 levels was valuated by means of a multivariate regression model. The results revealed that traffic low is the most important factor that controls PM2.5 hourly concentrations while street washing activities did not influence fine particle mass levels.

Spatial linear global instability analysis of the HIFiRE-5 elliptic cone model flow

The linear instability of the three-dimensional boundary-layer over the HIFiRE-5 flight test geometry, i.e. a rounded-tip 2:1 elliptic cone, at Mach 7, has been analyzed through spatial BiGlobal analysis, in a effort to understand transition and accurately predict local heat loads on next-generation ight vehicles. The results at an intermediate axial section of the cone, Re x = 8x10 5, show three different families of spatially amplied linear global modes, the attachment-line and cross- ow modes known from earlier analyses, and a new global mode, peaking in the vicinity of the minor axis of the cone, termed \center-line mode". We discover that a sequence of symmetric and anti-symmetric centerline modes exist and, for the basic ow at hand, are maximally amplied around F* = 130kHz. The wavenumbers and spatial distribution of amplitude functions of the centerline modes are documented

Comparaciones entre el factor "G" de Cattell y algunas pruebas de condición física en mis alumnos de 5º de EGB

Una de las cuestiones que más me ha interesado durante el tiempo que he estado cursando estudios en el Instituto Nacional de Educación Física de Madrid, y posteriormente en mi actividad educativa, es la posible relación entre la condición, la capacidad o el rendimiento físico o deportivo y la capacidad del rendimiento intelectual, así como las matizaciones posibles dentro de este amplísimo y complejo tema. Muchas veces he tenido que escuchar de compañeros en las tareas educativas, afirmaciones rotundas sin ninguna base científica acerca de la relación positiva o negativa, aunque casi siempre negativa, entre la capacidad física e intelectual, muy a menudo hechos en base a un sujeto que suspende prácticamente todas sus asignaturas y sin embargo obtiene sobresaliente en su calificación de Educación Física. No siendo este un caso muy frecuente, a estos educadores no se les ha ocurrido pensar en que la causa de esa diferencia tan destacada en el nivel de calificaciones, pueda ser la motivación y no la capacidad, tampoco piensan aun en el caso de que lo que hiciese la diferencia fuese una cuestión de capacidades, el que se produzca en algún caso aislado no implica ni que tenga que ocurrir siempre ni que a nivel general exista una relación negativa. La observación directa de los niños componentes de los equipos de baloncesto del colegio en el que imparto mis clases, me ha hecho ver que la mayor parte de ellos obtienen buenos rendimientos académicos, pocos mediocres, y muy pocos realmente bajos. Estos niños han sido seleccionados para formar parte de los equipos fundamentalmente por su buena condición física, lo cual podría hacer pensar en una posible relación, aunque no necesaria, entre condición física y rendimiento académico. Sin embargo gran cantidad de factores aparte de la inteligencia pueden influir en el mencionado rendimiento. Llegando el momento de plantear la realización de mi trabajo fin de carrera, en gran parte mediatizado por todo lo expuesto anteriormente, decidí centrarlo en la posible relación entre los diversos factores de la condición física, ésta misma y la inteligencia general. Teniendo en cuenta que estos son los factores más primarios y por consiguiente los que menos se van a ver influenciados por otros.