Statistical models for short-term wind power ramp forecasting

La predicción de energía eólica ha desempeñado en la última década un papel fundamental en el aprovechamiento de este recurso renovable, ya que permite reducir el impacto que tiene la naturaleza fluctuante del viento en la actividad de diversos agentes implicados en su integración, tales como el operador del sistema o los agentes del mercado eléctrico. Los altos niveles de penetración eólica alcanzados recientemente por algunos países han puesto de manifiesto la necesidad de mejorar las predicciones durante eventos en los que se experimenta una variación importante de la potencia generada por un parque o un conjunto de ellos en un tiempo relativamente corto (del orden de unas pocas horas). Estos eventos, conocidos como rampas, no tienen una única causa, ya que pueden estar motivados por procesos meteorológicos que se dan en muy diferentes escalas espacio-temporales, desde el paso de grandes frentes en la macroescala a procesos convectivos locales como tormentas. Además, el propio proceso de conversión del viento en energía eléctrica juega un papel relevante en la ocurrencia de rampas debido, entre otros factores, a la relación no lineal que impone la curva de potencia del aerogenerador, la desalineación de la máquina con respecto al viento y la interacción aerodinámica entre aerogeneradores. En este trabajo se aborda la aplicación de modelos estadísticos a la predicción de rampas a muy corto plazo. Además, se investiga la relación de este tipo de eventos con procesos atmosféricos en la macroescala. Los modelos se emplean para generar predicciones de punto a partir del modelado estocástico de una serie temporal de potencia generada por un parque eólico. Los horizontes de predicción considerados van de una a seis horas. Como primer paso, se ha elaborado una metodología para caracterizar rampas en series temporales. La denominada función-rampa está basada en la transformada wavelet y proporciona un índice en cada paso temporal. Este índice caracteriza la intensidad de rampa en base a los gradientes de potencia experimentados en un rango determinado de escalas temporales. Se han implementado tres tipos de modelos predictivos de cara a evaluar el papel que juega la complejidad de un modelo en su desempeño: modelos lineales autorregresivos (AR), modelos de coeficientes variables (VCMs) y modelos basado en redes neuronales (ANNs). Los modelos se han entrenado en base a la minimización del error cuadrático medio y la configuración de cada uno de ellos se ha determinado mediante validación cruzada. De cara a analizar la contribución del estado macroescalar de la atmósfera en la predicción de rampas, se ha propuesto una metodología que permite extraer, a partir de las salidas de modelos meteorológicos, información relevante para explicar la ocurrencia de estos eventos. La metodología se basa en el análisis de componentes principales (PCA) para la síntesis de la datos de la atmósfera y en el uso de la información mutua (MI) para estimar la dependencia no lineal entre dos señales. Esta metodología se ha aplicado a datos de reanálisis generados con un modelo de circulación general (GCM) de cara a generar variables exógenas que posteriormente se han introducido en los modelos predictivos. Los casos de estudio considerados corresponden a dos parques eólicos ubicados en España. Los resultados muestran que el modelado de la serie de potencias permitió una mejora notable con respecto al modelo predictivo de referencia (la persistencia) y que al añadir información de la macroescala se obtuvieron mejoras adicionales del mismo orden. Estas mejoras resultaron mayores para el caso de rampas de bajada. Los resultados también indican distintos grados de conexión entre la macroescala y la ocurrencia de rampas en los dos parques considerados. Abstract One of the main drawbacks of wind energy is that it exhibits intermittent generation greatly depending on environmental conditions. Wind power forecasting has proven to be an effective tool for facilitating wind power integration from both the technical and the economical perspective. Indeed, system operators and energy traders benefit from the use of forecasting techniques, because the reduction of the inherent uncertainty of wind power allows them the adoption of optimal decisions. Wind power integration imposes new challenges as higher wind penetration levels are attained. Wind power ramp forecasting is an example of such a recent topic of interest. The term ramp makes reference to a large and rapid variation (1-4 hours) observed in the wind power output of a wind farm or portfolio. Ramp events can be motivated by a broad number of meteorological processes that occur at different time/spatial scales, from the passage of large-scale frontal systems to local processes such as thunderstorms and thermally-driven flows. Ramp events may also be conditioned by features related to the wind-to-power conversion process, such as yaw misalignment, the wind turbine shut-down and the aerodynamic interaction between wind turbines of a wind farm (wake effect). This work is devoted to wind power ramp forecasting, with special focus on the connection between the global scale and ramp events observed at the wind farm level. The framework of this study is the point-forecasting approach. Time series based models were implemented for very short-term prediction, this being characterised by prediction horizons up to six hours ahead. As a first step, a methodology to characterise ramps within a wind power time series was proposed. The so-called ramp function is based on the wavelet transform and it provides a continuous index related to the ramp intensity at each time step. The underlying idea is that ramps are characterised by high power output gradients evaluated under different time scales. A number of state-of-the-art time series based models were considered, namely linear autoregressive (AR) models, varying-coefficient models (VCMs) and artificial neural networks (ANNs). This allowed us to gain insights into how the complexity of the model contributes to the accuracy of the wind power time series modelling. The models were trained in base of a mean squared error criterion and the final set-up of each model was determined through cross-validation techniques. In order to investigate the contribution of the global scale into wind power ramp forecasting, a methodological proposal to identify features in atmospheric raw data that are relevant for explaining wind power ramp events was presented. The proposed methodology is based on two techniques: principal component analysis (PCA) for atmospheric data compression and mutual information (MI) for assessing non-linear dependence between variables. The methodology was applied to reanalysis data generated with a general circulation model (GCM). This allowed for the elaboration of explanatory variables meaningful for ramp forecasting that were utilized as exogenous variables by the forecasting models. The study covered two wind farms located in Spain. All the models outperformed the reference model (the persistence) during both ramp and non-ramp situations. Adding atmospheric information had a noticeable impact on the forecasting performance, specially during ramp-down events. Results also suggested different levels of connection between the ramp occurrence at the wind farm level and the global scale.

Linear latent force models using Gaussian Processes

Purely data-driven approaches for machine learning present difficulties when data are scarce relative to the complexity of the model or when the model is forced to extrapolate. On the other hand, purely mechanistic approaches need to identify and specify all the interactions in the problem at hand (which may not be feasible) and still leave the issue of how to parameterize the system. In this paper, we present a hybrid approach using Gaussian processes and differential equations to combine data-driven modeling with a physical model of the system. We show how different, physically inspired, kernel functions can be developed through sensible, simple, mechanistic assumptions about the underlying system. The versatility of our approach is illustrated with three case studies from motion capture, computational biology, and geostatistics.

Technology and characterization of GaN-HEMTs devices: high temperature and trapping effects

Los transistores de alta movilidad electrónica basados en GaN han sido objeto de una extensa investigación ya que tanto el GaN como sus aleaciones presentan unas excelentes propiedades eléctricas (alta movilidad, elevada concentración de portadores y campo eléctrico crítico alto). Aunque recientemente se han incluido en algunas aplicaciones comerciales, su expansión en el mercado está condicionada a la mejora de varios asuntos relacionados con su rendimiento y habilidad. Durante esta tesis se han abordado algunos de estos aspectos relevantes; por ejemplo, la fabricación de enhancement mode HEMTs, su funcionamiento a alta temperatura, el auto calentamiento y el atrapamiento de carga. Los HEMTs normalmente apagado o enhancement mode han atraído la atención de la comunidad científica dedicada al desarrollo de circuitos amplificadores y conmutadores de potencia, ya que su utilización disminuiría significativamente el consumo de potencia; además de requerir solamente una tensión de alimentación negativa, y reducir la complejidad del circuito y su coste. Durante esta tesis se han evaluado varias técnicas utilizadas para la fabricación de estos dispositivos: el ataque húmedo para conseguir el gate-recess en heterostructuras de InAl(Ga)N/GaN; y tratamientos basados en flúor (plasma CF4 e implantación de F) de la zona debajo de la puerta. Se han llevado a cabo ataques húmedos en heteroestructuras de InAl(Ga)N crecidas sobre sustratos de Si, SiC y zafiro. El ataque completo de la barrera se consiguió únicamente en las muestras con sustrato de Si. Por lo tanto, se puede deducir que la velocidad de ataque depende de la densidad de dislocaciones presentes en la estructura, ya que el Si presenta un peor ajuste del parámetro de red con el GaN. En relación a los tratamientos basados en flúor, se ha comprobado que es necesario realizar un recocido térmico después de la fabricación de la puerta para recuperar la heteroestructura de los daños causados durante dichos tratamientos. Además, el estudio de la evolución de la tensión umbral con el tiempo de recocido ha demostrado que en los HEMTs tratados con plasma ésta tiende a valores más negativos al aumentar el tiempo de recocido. Por el contrario, la tensión umbral de los HEMTs implantados se desplaza hacia valores más positivos, lo cual se atribuye a la introducción de iones de flúor a niveles más profundos de la heterostructura. Los transistores fabricados con plasma presentaron mejor funcionamiento en DC a temperatura ambiente que los implantados. Su estudio a alta temperatura ha revelado una reducción del funcionamiento de todos los dispositivos con la temperatura. Los valores iniciales de corriente de drenador y de transconductancia medidos a temperatura ambiente se recuperaron después del ciclo térmico, por lo que se deduce que dichos efectos térmicos son reversibles. Se han estudiado varios aspectos relacionados con el funcionamiento de los HEMTs a diferentes temperaturas. En primer lugar, se han evaluado las prestaciones de dispositivos de AlGaN/GaN sobre sustrato de Si con diferentes caps: GaN, in situ SiN e in situ SiN/GaN, desde 25 K hasta 550 K. Los transistores con in situ SiN presentaron los valores más altos de corriente drenador, transconductancia, y los valores más bajos de resistencia-ON, así como las mejores características en corte. Además, se ha confirmado que dichos dispositivos presentan gran robustez frente al estrés térmico. En segundo lugar, se ha estudiado el funcionamiento de transistores de InAlN/GaN con diferentes diseños y geometrías. Dichos dispositivos presentaron una reducción casi lineal de los parámetros en DC en el rango de temperaturas de 25°C hasta 225°C. Esto se debe principalmente a la dependencia térmica de la movilidad electrónica, y también a la reducción de la drift velocity con la temperatura. Además, los transistores con mayores longitudes de puerta mostraron una mayor reducción de su funcionamiento, lo cual se atribuye a que la drift velocity disminuye más considerablemente con la temperatura cuando el campo eléctrico es pequeño. De manera similar, al aumentar la distancia entre la puerta y el drenador, el funcionamiento del HEMT presentó una mayor reducción con la temperatura. Por lo tanto, se puede deducir que la degradación del funcionamiento de los HEMTs causada por el aumento de la temperatura depende tanto de la longitud de la puerta como de la distancia entre la puerta y el drenador. Por otra parte, la alta densidad de potencia generada en la región activa de estos transistores conlleva el auto calentamiento de los mismos por efecto Joule, lo cual puede degradar su funcionamiento y Habilidad. Durante esta tesis se ha desarrollado un simple método para la determinación de la temperatura del canal basado en medidas eléctricas. La aplicación de dicha técnica junto con la realización de simulaciones electrotérmicas han posibilitado el estudio de varios aspectos relacionados con el autocalentamiento. Por ejemplo, se han evaluado sus efectos en dispositivos sobre Si, SiC, y zafiro. Los transistores sobre SiC han mostrado menores efectos gracias a la mayor conductividad térmica del SiC, lo cual confirma el papel clave que desempeña el sustrato en el autocalentamiento. Se ha observado que la geometría del dispositivo tiene cierta influencia en dichos efectos, destacando que la distribución del calor generado en la zona del canal depende de la distancia entre la puerta y el drenador. Además, se ha demostrado que la temperatura ambiente tiene un considerable impacto en el autocalentamiento, lo que se atribuye principalmente a la dependencia térmica de la conductividad térmica de las capas y sustrato que forman la heterostructura. Por último, se han realizado numerosas medidas en pulsado para estudiar el atrapamiento de carga en HEMTs sobre sustratos de SiC con barreras de AlGaN y de InAlN. Los resultados obtenidos en los transistores con barrera de AlGaN han presentado una disminución de la corriente de drenador y de la transconductancia sin mostrar un cambio en la tensión umbral. Por lo tanto, se puede deducir que la posible localización de las trampas es la región de acceso entre la puerta y el drenador. Por el contrario, la reducción de la corriente de drenador observada en los dispositivos con barrera de InAlN llevaba asociado un cambio significativo en la tensión umbral, lo que implica la existencia de trampas situadas en la zona debajo de la puerta. Además, el significativo aumento del valor de la resistencia-ON y la degradación de la transconductancia revelan la presencia de trampas en la zona de acceso entre la puerta y el drenador. La evaluación de los efectos del atrapamiento de carga en dispositivos con diferentes geometrías ha demostrado que dichos efectos son menos notables en aquellos transistores con mayor longitud de puerta o mayor distancia entre puerta y drenador. Esta dependencia con la geometría se puede explicar considerando que la longitud y densidad de trampas de la puerta virtual son independientes de las dimensiones del dispositivo. Finalmente se puede deducir que para conseguir el diseño óptimo durante la fase de diseño no sólo hay que tener en cuenta la aplicación final sino también la influencia que tiene la geometría en los diferentes aspectos estudiados (funcionamiento a alta temperatura, autocalentamiento, y atrapamiento de carga). ABSTRACT GaN-based high electron mobility transistors have been under extensive research due to the excellent electrical properties of GaN and its related alloys (high carrier concentration, high mobility, and high critical electric field). Although these devices have been recently included in commercial applications, some performance and reliability issues need to be addressed for their expansion in the market. Some of these relevant aspects have been studied during this thesis; for instance, the fabrication of enhancement mode HEMTs, the device performance at high temperature, the self-heating and the charge trapping. Enhancement mode HEMTs have become more attractive mainly because their use leads to a significant reduction of the power consumption during the stand-by state. Moreover, they enable the fabrication of simpler power amplifier circuits and high-power switches because they allow the elimination of negativepolarity voltage supply, reducing significantly the circuit complexity and system cost. In this thesis, different techniques for the fabrication of these devices have been assessed: wet-etching for achieving the gate-recess in InAl(Ga)N/GaN devices and two different fluorine-based treatments (CF4 plasma and F implantation). Regarding the wet-etching, experiments have been carried out in InAl(Ga)N/GaN grown on different substrates: Si, sapphire, and SiC. The total recess of the barrier was achieved after 3 min of etching in devices grown on Si substrate. This suggests that the etch rate can critically depend on the dislocations present in the structure, since the Si exhibits the highest mismatch to GaN. Concerning the fluorine-based treatments, a post-gate thermal annealing was required to recover the damages caused to the structure during the fluorine-treatments. The study of the threshold voltage as a function of this annealing time has revealed that in the case of the plasma-treated devices it become more negative with the time increase. On the contrary, the threshold voltage of implanted HEMTs showed a positive shift when the annealing time was increased, which is attributed to the deep F implantation profile. Plasma-treated HEMTs have exhibited better DC performance at room temperature than the implanted devices. Their study at high temperature has revealed that their performance decreases with temperature. The initial performance measured at room temperature was recovered after the thermal cycle regardless of the fluorine treatment; therefore, the thermal effects were reversible. Thermal issues related to the device performance at different temperature have been addressed. Firstly, AlGaN/GaN HEMTs grown on Si substrate with different cap layers: GaN, in situ SiN, or in situ SiN/GaN, have been assessed from 25 K to 550 K. In situ SiN cap layer has been demonstrated to improve the device performance since HEMTs with this cap layer have exhibited the highest drain current and transconductance values, the lowest on-resistance, as well as the best off-state characteristics. Moreover, the evaluation of thermal stress impact on the device performance has confirmed the robustness of devices with in situ cap. Secondly, the high temperature performance of InAlN/GaN HEMTs with different layouts and geometries have been assessed. The devices under study have exhibited an almost linear reduction of the main DC parameters operating in a temperature range from room temperature to 225°C. This was mainly due to the thermal dependence of the electron mobility, and secondly to the drift velocity decrease with temperature. Moreover, HEMTs with large gate length values have exhibited a great reduction of the device performance. This was attributed to the greater decrease of the drift velocity for low electric fields. Similarly, the increase of the gate-to-drain distance led to a greater reduction of drain current and transconductance values. Therefore, this thermal performance degradation has been found to be dependent on both the gate length and the gate-to-drain distance. It was observed that the very high power density in the active region of these transistors leads to Joule self-heating, resulting in an increase of the device temperature, which can degrade the device performance and reliability. A simple electrical method have been developed during this work to determine the channel temperature. Furthermore, the application of this technique together with the performance of electro-thermal simulations have enabled the evaluation of different aspects related to the self-heating. For instance, the influence of the substrate have been confirmed by the study of devices grown on Si, SiC, and Sapphire. HEMTs grown on SiC substrate have been confirmed to exhibit the lowest self-heating effects thanks to its highest thermal conductivity. In addition to this, the distribution of the generated heat in the channel has been demonstrated to be dependent on the gate-to-drain distance. Besides the substrate and the geometry of the device, the ambient temperature has also been found to be relevant for the self-heating effects, mainly due to the temperature-dependent thermal conductivity of the layers and the substrate. Trapping effects have been evaluated by means of pulsed measurements in AlGaN and InAIN barrier devices. AlGaN barrier HEMTs have exhibited a de crease in drain current and transconductance without measurable threshold voltage change, suggesting the location of the traps in the gate-to-drain access region. On the contrary, InAIN barrier devices have showed a drain current associated with a positive shift of threshold voltage, which indicated that the traps were possibly located under the gate region. Moreover, a significant increase of the ON-resistance as well as a transconductance reduction were observed, revealing the presence of traps on the gate-drain access region. On the other hand, the assessment of devices with different geometries have demonstrated that the trapping effects are more noticeable in devices with either short gate length or the gate-to-drain distance. This can be attributed to the fact that the length and the trap density of the virtual gate are independent on the device geometry. Finally, it can be deduced that besides the final application requirements, the influence of the device geometry on the performance at high temperature, on the self-heating, as well as on the trapping effects need to be taken into account during the device design stage to achieve the optimal layout.

Non uniform embedding based on relevance analysis with reduced computational complexity: application to the detection of pathologies from biosignal recordings

Nonlinear analysis tools for studying and characterizing the dynamics of physiological signals have gained popularity, mainly because tracking sudden alterations of the inherent complexity of biological processes might be an indicator of altered physiological states. Typically, in order to perform an analysis with such tools, the physiological variables that describe the biological process under study are used to reconstruct the underlying dynamics of the biological processes. For that goal, a procedure called time-delay or uniform embedding is usually employed. Nonetheless, there is evidence of its inability for dealing with non-stationary signals, as those recorded from many physiological processes. To handle with such a drawback, this paper evaluates the utility of non-conventional time series reconstruction procedures based on non uniform embedding, applying them to automatic pattern recognition tasks. The paper compares a state of the art non uniform approach with a novel scheme which fuses embedding and feature selection at once, searching for better reconstructions of the dynamics of the system. Moreover, results are also compared with two classic uniform embedding techniques. Thus, the goal is comparing uniform and non uniform reconstruction techniques, including the one proposed in this work, for pattern recognition in biomedical signal processing tasks. Once the state space is reconstructed, the scheme followed characterizes with three classic nonlinear dynamic features (Largest Lyapunov Exponent, Correlation Dimension and Recurrence Period Density Entropy), while classification is carried out by means of a simple k-nn classifier. In order to test its generalization capabilities, the approach was tested with three different physiological databases (Speech Pathologies, Epilepsy and Heart Murmurs). In terms of the accuracy obtained to automatically detect the presence of pathologies, and for the three types of biosignals analyzed, the non uniform techniques used in this work lightly outperformed the results obtained using the uniform methods, suggesting their usefulness to characterize non-stationary biomedical signals in pattern recognition applications. On the other hand, in view of the results obtained and its low computational load, the proposed technique suggests its applicability for the applications under study.

Logarithmical hopping encoding: a low computational complexity algorithm for image compression

LHE (logarithmical hopping encoding) is a computationally efficient image compression algorithm that exploits the Weber–Fechner law to encode the error between colour component predictions and the actual value of such components. More concretely, for each pixel, luminance and chrominance predictions are calculated as a function of the surrounding pixels and then the error between the predictions and the actual values are logarithmically quantised. The main advantage of LHE is that although it is capable of achieving a low-bit rate encoding with high quality results in terms of peak signal-to-noise ratio (PSNR) and image quality metrics with full-reference (FSIM) and non-reference (blind/referenceless image spatial quality evaluator), its time complexity is O( n) and its memory complexity is O(1). Furthermore, an enhanced version of the algorithm is proposed, where the output codes provided by the logarithmical quantiser are used in a pre-processing stage to estimate the perceptual relevance of the image blocks. This allows the algorithm to downsample the blocks with low perceptual relevance, thus improving the compression rate. The performance of LHE is especially remarkable when the bit per pixel rate is low, showing much better quality, in terms of PSNR and FSIM, than JPEG and slightly lower quality than JPEG-2000 but being more computationally efficient.

Optimal Piecewise Linear Function Approximation for GPU-based Applications

Many computer vision and human-computer interaction applications developed in recent years need evaluating complex and continuous mathematical functions as an essential step toward proper operation. However, rigorous evaluation of this kind of functions often implies a very high computational cost, unacceptable in real-time applications. To alleviate this problem, functions are commonly approximated by simpler piecewise-polynomial representations. Following this idea, we propose a novel, efficient, and practical technique to evaluate complex and continuous functions using a nearly optimal design of two types of piecewise linear approximations in the case of a large budget of evaluation subintervals. To this end, we develop a thorough error analysis that yields asymptotically tight bounds to accurately quantify the approximation performance of both representations. It provides an improvement upon previous error estimates and allows the user to control the trade-off between the approximation error and the number of evaluation subintervals. To guarantee real-time operation, the method is suitable for, but not limited to, an efficient implementation in modern Graphics Processing Units (GPUs), where it outperforms previous alternative approaches by exploiting the fixed-function interpolation routines present in their texture units. The proposed technique is a perfect match for any application requiring the evaluation of continuous functions, we have measured in detail its quality and efficiency on several functions, and, in particular, the Gaussian function because it is extensively used in many areas of computer vision and cybernetics, and it is expensive to evaluate.

Age regression from soft aligned face images using low computational resources

The initial step in most facial age estimation systems consists of accurately aligning a model to the output of a face detector (e.g. an Active Appearance Model). This fitting process is very expensive in terms of computational resources and prone to get stuck in local minima. This makes it impractical for analysing faces in resource limited computing devices. In this paper we build a face age regressor that is able to work directly on faces cropped using a state-of-the-art face detector. Our procedure uses K nearest neighbours (K-NN) regression with a metric based on a properly tuned Fisher Linear Discriminant Analysis (LDA) projection matrix. On FG-NET we achieve a state-of-the-art Mean Absolute Error (MAE) of 5.72 years with manually aligned faces. Using face images cropped by a face detector we get a MAE of 6.87 years in the same database. Moreover, most of the algorithms presented in the literature have been evaluated on single database experiments and therefore, they report optimistically biased results. In our cross-database experiments we get a MAE of roughly 12 years, which would be the expected performance in a real world application.

Estimación eficiente de atributos demográficos del rostro humano en imágenes

Sin duda, el rostro humano ofrece mucha más información de la que pensamos. La cara transmite sin nuestro consentimiento señales no verbales, a partir de las interacciones faciales, que dejan al descubierto nuestro estado afectivo, actividad cognitiva, personalidad y enfermedades. Estudios recientes [OFT14, TODMS15] demuestran que muchas de nuestras decisiones sociales e interpersonales derivan de un previo análisis facial de la cara que nos permite establecer si esa persona es confiable, trabajadora, inteligente, etc. Esta interpretación, propensa a errores, deriva de la capacidad innata de los seres humanas de encontrar estas señales e interpretarlas. Esta capacidad es motivo de estudio, con un especial interés en desarrollar métodos que tengan la habilidad de calcular de manera automática estas señales o atributos asociados a la cara. Así, el interés por la estimación de atributos faciales ha crecido rápidamente en los últimos años por las diversas aplicaciones en que estos métodos pueden ser utilizados: marketing dirigido, sistemas de seguridad, interacción hombre-máquina, etc. Sin embargo, éstos están lejos de ser perfectos y robustos en cualquier dominio de problemas. La principal dificultad encontrada es causada por la alta variabilidad intra-clase debida a los cambios en la condición de la imagen: cambios de iluminación, oclusiones, expresiones faciales, edad, género, etnia, etc.; encontradas frecuentemente en imágenes adquiridas en entornos no controlados. Este de trabajo de investigación estudia técnicas de análisis de imágenes para estimar atributos faciales como el género, la edad y la postura, empleando métodos lineales y explotando las dependencias estadísticas entre estos atributos. Adicionalmente, nuestra propuesta se centrará en la construcción de estimadores que tengan una fuerte relación entre rendimiento y coste computacional. Con respecto a éste último punto, estudiamos un conjunto de estrategias para la clasificación de género y las comparamos con una propuesta basada en un clasificador Bayesiano y una adecuada extracción de características. Analizamos en profundidad el motivo de porqué las técnicas lineales no han logrado resultados competitivos hasta la fecha y mostramos cómo obtener rendimientos similares a las mejores técnicas no-lineales. Se propone un segundo algoritmo para la estimación de edad, basado en un regresor K-NN y una adecuada selección de características tal como se propuso para la clasificación de género. A partir de los experimentos desarrollados, observamos que el rendimiento de los clasificadores se reduce significativamente si los ´estos han sido entrenados y probados sobre diferentes bases de datos. Hemos encontrado que una de las causas es la existencia de dependencias entre atributos faciales que no han sido consideradas en la construcción de los clasificadores. Nuestro resultados demuestran que la variabilidad intra-clase puede ser reducida cuando se consideran las dependencias estadísticas entre los atributos faciales de el género, la edad y la pose; mejorando el rendimiento de nuestros clasificadores de atributos faciales con un coste computacional pequeño. Abstract Surely the human face provides much more information than we think. The face provides without our consent nonverbal cues from facial interactions that reveal our emotional state, cognitive activity, personality and disease. Recent studies [OFT14, TODMS15] show that many of our social and interpersonal decisions derive from a previous facial analysis that allows us to establish whether that person is trustworthy, hardworking, intelligent, etc. This error-prone interpretation derives from the innate ability of human beings to find and interpret these signals. This capability is being studied, with a special interest in developing methods that have the ability to automatically calculate these signs or attributes associated with the face. Thus, the interest in the estimation of facial attributes has grown rapidly in recent years by the various applications in which these methods can be used: targeted marketing, security systems, human-computer interaction, etc. However, these are far from being perfect and robust in any domain of problems. The main difficulty encountered is caused by the high intra-class variability due to changes in the condition of the image: lighting changes, occlusions, facial expressions, age, gender, ethnicity, etc.; often found in images acquired in uncontrolled environments. This research work studies image analysis techniques to estimate facial attributes such as gender, age and pose, using linear methods, and exploiting the statistical dependencies between these attributes. In addition, our proposal will focus on the construction of classifiers that have a good balance between performance and computational cost. We studied a set of strategies for gender classification and we compare them with a proposal based on a Bayesian classifier and a suitable feature extraction based on Linear Discriminant Analysis. We study in depth why linear techniques have failed to provide competitive results to date and show how to obtain similar performances to the best non-linear techniques. A second algorithm is proposed for estimating age, which is based on a K-NN regressor and proper selection of features such as those proposed for the classification of gender. From our experiments we note that performance estimates are significantly reduced if they have been trained and tested on different databases. We have found that one of the causes is the existence of dependencies between facial features that have not been considered in the construction of classifiers. Our results demonstrate that intra-class variability can be reduced when considering the statistical dependencies between facial attributes gender, age and pose, thus improving the performance of our classifiers with a reduced computational cost.

Parallel Computer Vision Algorithms for Graphics Processing Units

La evolución de los teléfonos móviles inteligentes, dotados de cámaras digitales, está provocando una creciente demanda de aplicaciones cada vez más complejas que necesitan algoritmos de visión artificial en tiempo real; puesto que el tamaño de las señales de vídeo no hace sino aumentar y en cambio el rendimiento de los procesadores de un solo núcleo se ha estancado, los nuevos algoritmos que se diseñen para visión artificial han de ser paralelos para poder ejecutarse en múltiples procesadores y ser computacionalmente escalables. Una de las clases de procesadores más interesantes en la actualidad se encuentra en las tarjetas gráficas (GPU), que son dispositivos que ofrecen un alto grado de paralelismo, un excelente rendimiento numérico y una creciente versatilidad, lo que los hace interesantes para llevar a cabo computación científica. En esta tesis se exploran dos aplicaciones de visión artificial que revisten una gran complejidad computacional y no pueden ser ejecutadas en tiempo real empleando procesadores tradicionales. En cambio, como se demuestra en esta tesis, la paralelización de las distintas subtareas y su implementación sobre una GPU arrojan los resultados deseados de ejecución con tasas de refresco interactivas. Asimismo, se propone una técnica para la evaluación rápida de funciones de complejidad arbitraria especialmente indicada para su uso en una GPU. En primer lugar se estudia la aplicación de técnicas de síntesis de imágenes virtuales a partir de únicamente dos cámaras lejanas y no paralelas—en contraste con la configuración habitual en TV 3D de cámaras cercanas y paralelas—con información de color y profundidad. Empleando filtros de mediana modificados para la elaboración de un mapa de profundidad virtual y proyecciones inversas, se comprueba que estas técnicas son adecuadas para una libre elección del punto de vista. Además, se demuestra que la codificación de la información de profundidad con respecto a un sistema de referencia global es sumamente perjudicial y debería ser evitada. Por otro lado se propone un sistema de detección de objetos móviles basado en técnicas de estimación de densidad con funciones locales. Este tipo de técnicas es muy adecuada para el modelado de escenas complejas con fondos multimodales, pero ha recibido poco uso debido a su gran complejidad computacional. El sistema propuesto, implementado en tiempo real sobre una GPU, incluye propuestas para la estimación dinámica de los anchos de banda de las funciones locales, actualización selectiva del modelo de fondo, actualización de la posición de las muestras de referencia del modelo de primer plano empleando un filtro de partículas multirregión y selección automática de regiones de interés para reducir el coste computacional. Los resultados, evaluados sobre diversas bases de datos y comparados con otros algoritmos del estado del arte, demuestran la gran versatilidad y calidad de la propuesta. Finalmente se propone un método para la aproximación de funciones arbitrarias empleando funciones continuas lineales a tramos, especialmente indicada para su implementación en una GPU mediante el uso de las unidades de filtraje de texturas, normalmente no utilizadas para cómputo numérico. La propuesta incluye un riguroso análisis matemático del error cometido en la aproximación en función del número de muestras empleadas, así como un método para la obtención de una partición cuasióptima del dominio de la función para minimizar el error. ABSTRACT The evolution of smartphones, all equipped with digital cameras, is driving a growing demand for ever more complex applications that need to rely on real-time computer vision algorithms. However, video signals are only increasing in size, whereas the performance of single-core processors has somewhat stagnated in the past few years. Consequently, new computer vision algorithms will need to be parallel to run on multiple processors and be computationally scalable. One of the most promising classes of processors nowadays can be found in graphics processing units (GPU). These are devices offering a high parallelism degree, excellent numerical performance and increasing versatility, which makes them interesting to run scientific computations. In this thesis, we explore two computer vision applications with a high computational complexity that precludes them from running in real time on traditional uniprocessors. However, we show that by parallelizing subtasks and implementing them on a GPU, both applications attain their goals of running at interactive frame rates. In addition, we propose a technique for fast evaluation of arbitrarily complex functions, specially designed for GPU implementation. First, we explore the application of depth-image–based rendering techniques to the unusual configuration of two convergent, wide baseline cameras, in contrast to the usual configuration used in 3D TV, which are narrow baseline, parallel cameras. By using a backward mapping approach with a depth inpainting scheme based on median filters, we show that these techniques are adequate for free viewpoint video applications. In addition, we show that referring depth information to a global reference system is ill-advised and should be avoided. Then, we propose a background subtraction system based on kernel density estimation techniques. These techniques are very adequate for modelling complex scenes featuring multimodal backgrounds, but have not been so popular due to their huge computational and memory complexity. The proposed system, implemented in real time on a GPU, features novel proposals for dynamic kernel bandwidth estimation for the background model, selective update of the background model, update of the position of reference samples of the foreground model using a multi-region particle filter, and automatic selection of regions of interest to reduce computational cost. The results, evaluated on several databases and compared to other state-of-the-art algorithms, demonstrate the high quality and versatility of our proposal. Finally, we propose a general method for the approximation of arbitrarily complex functions using continuous piecewise linear functions, specially formulated for GPU implementation by leveraging their texture filtering units, normally unused for numerical computation. Our proposal features a rigorous mathematical analysis of the approximation error in function of the number of samples, as well as a method to obtain a suboptimal partition of the domain of the function to minimize approximation error.

The conducting form of gramicidin A is a right-handed double-stranded double helix

The linear pentadecapeptide antibiotic, gramicidin D, is a naturally occurring product of Bacillus brevis known to form ion channels in synthetic and natural membranes. The x-ray crystal structures of the right-handed double-stranded double-helical dimers (DSDHℛ) reported here agree with 15N-NMR and CD data on the functional gramicidin D channel in lipid bilayers. These structures demonstrate single-file ion transfer through the channels. The results also indicate that previous crystal structure reports of a left-handed double-stranded double-helical dimer in complex with Cs+ and K+ salts may be in error and that our evidence points to the DSDHℛ as the major conformer responsible for ion transport in membranes.

Hybridization of single-stranded DNA targets to immobilized complementary DNA probes: comparison of hairpin versus linear capture probes

A microtiter-based assay system is described in which DNA hairpin probes with dangling ends and single-stranded, linear DNA probes were immobilized and compared based on their ability to capture single-strand target DNA. Hairpin probes consisted of a 16 bp duplex stem, linked by a T2-biotin·dT-T2 loop. The third base was a biotinylated uracil (UB) necessary for coupling to avidin coated microtiter wells. The capture region of the hairpin was a 3′ dangling end composed of either 16 or 32 bases. Fundamental parameters of the system, such as probe density and avidin adsorption capacity of the plates were characterized. The target DNA consisted of 65 bases whose 3′ end was complementary to the dangling end of the hairpin or to the linear probe sequence. The assay system was employed to measure the time dependence and thermodynamic stability of target hybridization with hairpin and linear probes. Target molecules were labeled with either a 5′-FITC, or radiolabeled with [γ-33P]ATP and captured by either linear or hairpin probes affixed to the solid support. Over the range of target concentrations from 10 to 640 pmol hybridization rates increased with increasing target concentration, but varied for the different probes examined. Hairpin probes displayed higher rates of hybridization and larger equilibrium amounts of captured targets than linear probes. At 25 and 45°C, rates of hybridization were better than twice as great for the hairpin compared with the linear capture probes. Hairpin–target complexes were also more thermodynamically stable. Binding free energies were evaluated from the observed equilibrium constants for complex formation. Results showed the order of stability of the probes to be: hairpins with 32 base dangling ends > hairpin probes with l6 base dangling ends > 16 base linear probes > 32 base linear probes. The physical characteristics of hairpins could offer substantial advantages as nucleic acid capture moieties in solid support based hybridization systems.

Complexity of the erythroid transcription factor NF-E2 as revealed by gene targeting of the mouse p18 NF-E2 locus.

High-level globin expression in erythroid precursor cells depends on the integrity of NF-E2 recognition sites, transcription factor AP-1-like protein-binding motifs, located in the upstream regulatory regions of the alpha- and beta-globin loci. The NF-E2 transcription factor, which recognizes these sites, is a heterodimer consisting of (i) p45 NF-E2 (the larger subunit), a hematopoietic-restricted basic leucine zipper protein, and (ii) a widely expressed basic leucine zipper factor, p18 NF-E2, the smaller subunit. p18 NF-E2 protein shares extensive homology with the maf protooncogene family. To determine an in vivo role for p18 NF-E2 protein we disrupted the p18 NF-E2-encoding gene by homologous recombination in murine embryonic stem cells and generated p18 NF-E2-/- mice. These mice are indistinguishable from littermates throughout all phases of development and remain healthy in adulthood. Despite the absence of expressed p18 NF-E2, DNA-binding activity with the properties of the NF-E2 heterodimer is present in fetal liver erythroid cells of p18 NF-E2-/- mice. We speculate that another member of the maf basic leucine zipper family substitutes for the p18 subunit in a complex with p45 NF-E2. Thus, p18 NF-E2 per se appears to be dispensable in vivo.

Impact of artifact correction methods on R-R interbeat signals to quantifying heart rate variability (HRV) according to linear and nonlinear methods.

In the analysis of heart rate variability (HRV) are used temporal series that contains the distances between successive heartbeats in order to assess autonomic regulation of the cardiovascular system. These series are obtained from the electrocardiogram (ECG) signal analysis, which can be affected by different types of artifacts leading to incorrect interpretations in the analysis of the HRV signals. Classic approach to deal with these artifacts implies the use of correction methods, some of them based on interpolation, substitution or statistical techniques. However, there are few studies that shows the accuracy and performance of these correction methods on real HRV signals. This study aims to determine the performance of some linear and non-linear correction methods on HRV signals with induced artefacts by quantification of its linear and nonlinear HRV parameters. As part of the methodology, ECG signals of rats measured using the technique of telemetry were used to generate real heart rate variability signals without any error. In these series were simulated missing points (beats) in different quantities in order to emulate a real experimental situation as accurately as possible. In order to compare recovering efficiency, deletion (DEL), linear interpolation (LI), cubic spline interpolation (CI), moving average window (MAW) and nonlinear predictive interpolation (NPI) were used as correction methods for the series with induced artifacts. The accuracy of each correction method was known through the results obtained after the measurement of the mean value of the series (AVNN), standard deviation (SDNN), root mean square error of the differences between successive heartbeats (RMSSD), Lomb\'s periodogram (LSP), Detrended Fluctuation Analysis (DFA), multiscale entropy (MSE) and symbolic dynamics (SD) on each HRV signal with and without artifacts. The results show that, at low levels of missing points the performance of all correction techniques are very similar with very close values for each HRV parameter. However, at higher levels of losses only the NPI method allows to obtain HRV parameters with low error values and low quantity of significant differences in comparison to the values calculated for the same signals without the presence of missing points.

Examining the impact of pre-k and kindergarten classroom alignment on school readiness outcomes

This study explored children’s experiences of instructional alignment from prekindergarten to kindergarten and analyzed the impact of those alignment experiences on children’s school readiness outcomes. The study answered the following overarching research question: Does the alignment of children’s learning experiences between prekindergarten and kindergarten impact school readiness outcomes? Three sub-questions drove the research design: (1) How do children’s prekindergarten and kindergarten learning experiences align; (2) To what extent does the alignment of early learning experiences predict children’s school readiness outcomes; and (3) Does the quality of prekindergarten classroom teacher interactions moderate the impact of any PK-K alignment effects? Using cluster analysis and hierarchical linear modeling (HLM) to analyze data from over 1,300 children in the 2009 Head Start Family and Child Experiences Survey (FACES), the study found that children have distinct and definable experiences of PK-K alignment. Results also indicated a disparity in children’s PK-K alignment experiences, with Hispanic/Latino children more likely to attend Head Start programs with poor systems transition practices followed by kindergartens with poor classroom structures. The study found that growth in the use of instructional activity centers from prekindergarten to kindergarten is predictive of better literacy and math outcomes. Findings further suggested that boys, minority students, and children from lower income households are predicted to score lower than girls, white classmates, and higher-income peers across school readiness measures. Findings support the need for equitable transition and alignment practices for children from all racial and ethnic groups. They also argue for an increase in child-directed activity centers in kindergarten. With one exception, the current findings did not support the hypothesis that prekindergarten teacher quality is a moderator of alignment effects on children’s school readiness outcomes. The study presents suggestions for further research.

Corn stover thermal decomposition in pyrolytic and oxidant atmosphere

The pyrolysis and combustion of corn stover were studied by dynamic thermogravimetry and derivate thermogravimetry (TG-DTG) at heating rates of 5, 10, 20 and 50 K min−1 at atmospheric pressure. For the simulation of pyrolysis and combustion processes a kinetic model based on the distribution of activation energies was used, with three pools of reactants (three pseudocomponents) because of the complexity of the biomass samples of agricultural origin. The experimental thermogravimetric data of pyrolysis and combustion processes were simultaneously fitted to determine a single set of kinetic parameters able to describe both processes at the different heating rates. The model proposed achieves a good correlation between the experimental and calculated curves, with an error of less than 4% for fitting four heating rates simultaneously. The experimental results and kinetic parameters may provide useful data for the design of thermo decomposition processing system using corn stover as feedstock. On the other hand, analysis of the main compounds in the evolved gas is given by means of a microcromatograph.