User-definable resource bounds analysis for logic programs

We present a static analysis that infers both upper and lower bounds on the usage that a logic program makes of a set of user-definable resources. The inferred bounds will in general be functions of input data sizes. A resource in our approach is a quite general, user-defined notion which associates a basic cost function with elementary operations. The analysis then derives the related (upper- and lower-bound) resource usage functions for all predicates in the program. We also present an assertion language which is used to define both such resources and resourcerelated properties that the system can then check based on the results of the analysis. We have performed some preliminary experiments with some concrete resources such as execution steps, bytes sent or received by an application, number of files left open, number of accesses to a datábase, number of calis to a procedure, number of asserts/retracts, etc. Applications of our analysis include resource consumption verification and debugging (including for mobile code), resource control in parallel/distributed computing, and resource-oriented specialization.

Resource bounds analysis.

We present a generic analysis that infers both upper and lower bounds on the usage that a program makes of a set of user-definable resources. The inferred bounds will in general be functions of input data sizes. A resource in our approach is a quite general, user-defined notion which associates a basic cost function with elementary operations. The analysis then derives the related (upper- and lower- bound) cost functions for all procedures in the program. We also present an assertion language which is used to define both such resources and resource-related properties that the system can then check based on the results of the analysis. We have performed some experiments with some concrete resource-related properties such as execution steps, bits sent or received by an application, number of arithmetic operations performed, number of calls to a procedure, number of transactions, etc. presenting the resource usage functions inferred and the times taken to perform the analysis. Applications of our analysis include resource consumption verification and debugging (including for mobile code), resource control in parallel/distributed computing, and resource-oriented specialization.

Distributes award protocol: A cooperation and communication method for robots.

In this Master’s Thesis a new Distributed Award Protocol (DAP) for robot communication and cooperation is presented. Task assignment (contract awarding) is done dynamically with contracts assigned to robots based upon the best bid received. Instead of having a manager and a contractor it is proposed a fully distributed bidding/awarding mechanism without a distinguished master. The best bidding robots are awarded with contract for execution. The contractors make decisions locally. This brings the following benefits: no communication bottleneck, low computational power requirement, increased robustness. DAP can handle multitasking. Tasks can be injected into system during the execution of already allocated tasks. As tasks have priorities, in the next cycle after taking into account actual bid parameters of all robots, tasks can be re-allocated. The aim is to minimize a global cost function which is a compromise between cost of task execution and cost of resources usage. Information about tasks and bid values is spread among robots with the use of a Round Robin Route, which is a novel solution proposed in this work. This method allows also identifying failed robots. Such failed robot is eliminated from the list of awarded robots and its replacement is found so the task is still executed by a team. If the failure of a robot was temporary (e.g. communication noise) and the robot can recover, it can again participate in the next bidding/awarding process. Using a bidding/awarding mechanism allows robots to dynamically relocate among tasks. This is also contributes to system robustness. DAP was evaluated through multiple experiments done in the multi-robot simulation system. Various scenarios were tested to check the idea of the main algorithm. Different failures of robots (communication failures, partial hardware malfunctions) were simulated and observations were made regarding how DAP recovers from them. Also the DAP flexibility to environment changes was watched. The experiments in the simulated environment confirmed the above features of DAP.

Cross-Products LASSO

Negative co-occurrence is a common phenomenon in many signal processing applications. In some cases the signals involved are sparse, and this information can be exploited to recover them. In this paper, we present a sparse learning approach that explicitly takes into account negative co-occurrence. This is achieved by adding a novel penalty term to the LASSO cost function based on the cross-products between the reconstruction coefficients. Although the resulting optimization problem is non-convex, we develop a new and efficient method for solving it based on successive convex approximations. Results on synthetic data, for both complete and overcomplete dictionaries, are provided to validate the proposed approach.

NURBS-Based Geometry Parameterization for Aerodynamic Shape Optimization

La motivación de esta tesis es el desarrollo de una herramienta de optimización automática para la mejora del rendimiento de formas aerodinámicas enfocado en la industria aeronáutica. Este trabajo cubre varios aspectos esenciales, desde el empleo de Non-Uniform Rational B-Splines (NURBS), al cálculo de gradientes utilizando la metodología del adjunto continuo, el uso de b-splines volumétricas como parámetros de diseño, el tratamiento de la malla en las intersecciones, y no menos importante, la adaptación de los algoritmos de la dinámica de fluidos computacional (CFD) en arquitecturas hardware de alto paralelismo, como las tarjetas gráficas, para acelerar el proceso de optimización. La metodología adjunta ha posibilitado que los métodos de optimización basados en gradientes sean una alternativa prometedora para la mejora de la eficiencia aerodinámica de los aviones. La formulación del adjunto permite calcular los gradientes de una función de coste, como la resistencia aerodinámica o la sustentación, independientemente del número de variables de diseño, a un coste computacional equivalente a una simulación CFD. Sin embargo, existen problemas prácticos que han imposibilitado su aplicación en la industria, que se pueden resumir en: integrabilidad, rendimiento computacional y robustez de la solución adjunta. Este trabajo aborda estas contrariedades y las analiza en casos prácticos. Como resumen, las contribuciones de esta tesis son: • El uso de NURBS como variables de diseño en un bucle de automático de optimización, aplicado a la mejora del rendimiento aerodinámico de alas en régimen transónico. • El desarrollo de algoritmos de inversión de punto, para calcular las coordenadas paramétricas de las coordenadas espaciales, para ligar los vértices de malla a las NURBS. • El uso y validación de la formulación adjunta para el calculo de los gradientes, a partir de las sensibilidades de la solución adjunta, comparado con diferencias finitas. • Se ofrece una estrategia para utilizar la geometría CAD, en forma de parches NURBS, para tratar las intersecciones, como el ala-fuselaje. • No existen muchas alternativas de librerías NURBS viables. En este trabajo se ha desarrollado una librería, DOMINO NURBS, y se ofrece a la comunidad como código libre y abierto. • También se ha implementado un código CFD en tarjeta gráfica, para realizar una valoración de cómo se puede adaptar un código sobre malla no estructurada a arquitecturas paralelas. • Finalmente, se propone una metodología, basada en la función de Green, como una forma eficiente de paralelizar simulaciones numéricas. Esta tesis ha sido apoyada por las actividades realizadas por el Área de Dinámica da Fluidos del Instituto Nacional de Técnica Aeroespacial (INTA), a través de numerosos proyectos de financiación nacional: DOMINO, SIMUMAT, y CORESFMULAERO. También ha estado en consonancia con las actividades realizadas por el departamento de Métodos y Herramientas de Airbus España y con el grupo Investigación y Tecnología Aeronáutica Europeo (GARTEUR), AG/52. ABSTRACT The motivation of this work is the development of an automatic optimization strategy for large scale shape optimization problems that arise in the aeronautics industry to improve the aerodynamic performance; covering several aspects from the use of Non-Uniform Rational B-Splines (NURBS), the calculation of the gradients with the continuous adjoint formulation, the development of volumetric b-splines parameterization, mesh adaptation and intersection handling, to the adaptation of Computational Fluid Dynamics (CFD) algorithms to take advantage of highly parallel architectures in order to speed up the optimization process. With the development of the adjoint formulation, gradient-based methods for aerodynamic optimization become a promising approach to improve the aerodynamic performance of aircraft designs. The adjoint methodology allows the evaluation the gradients to all design variables of a cost function, such as drag or lift, at the equivalent cost of more or less one CFD simulation. However, some practical problems have been delaying its full implementation to the industry, which can be summarized as: integrability, computer performance, and adjoint robustness. This work tackles some of these issues and analyse them in well-known test cases. As summary, the contributions comprises: • The employment of NURBS as design variables in an automatic optimization loop for the improvement of the aerodynamic performance of aircraft wings in transonic regimen. • The development of point inversion algorithms to calculate the NURBS parametric coordinates from the space coordinates, to link with the computational grid vertex. • The use and validation of the adjoint formulation to calculate the gradients from the surface sensitivities in an automatic optimization loop and evaluate its reliability, compared with finite differences. • This work proposes some algorithms that take advantage of the underlying CAD geometry description, in the form of NURBS patches, to handle intersections and mesh adaptations. • There are not many usable libraries for NURBS available. In this work an open source library DOMINO NURBS has been developed and is offered to the community as free, open source code. • The implementation of a transonic CFD solver from scratch in a graphic card, for an assessment of the implementability of conventional CFD solvers for unstructured grids to highly parallel architectures. • Finally, this research proposes the use of the Green's function as an efficient paralellization scheme of numerical solvers. The presented work has been supported by the activities carried out at the Fluid Dynamics branch of the National Institute for Aerospace Technology (INTA) through national founding research projects: DOMINO, SIMUMAT, and CORESIMULAERO; in line with the activities carried out by the Methods and Tools and Flight Physics department at Airbus and the Group for Aeronautical Research and Technology in Europe (GARTEUR) action group AG/52.

Modelo de gestión de un embalse en tiempo real durante avenidas basado en redes bayesianas entrenadas con el método de optimización PLEM

En la actualidad, la gestión de embalses para el control de avenidas se realiza, comúnmente, utilizando modelos de simulación. Esto se debe, principalmente, a su facilidad de uso en tiempo real por parte del operador de la presa. Se han desarrollado modelos de optimización de la gestión del embalse que, aunque mejoran los resultados de los modelos de simulación, su aplicación en tiempo real se hace muy difícil o simplemente inviable, pues está limitada al conocimiento de la avenida futura que entra al embalse antes de tomar la decisión de vertido. Por esta razón, se ha planteado el objetivo de desarrollar un modelo de gestión de embalses en avenidas que incorpore las ventajas de un modelo de optimización y que sea de fácil uso en tiempo real por parte del gestor de la presa. Para ello, se construyó un modelo de red Bayesiana que representa los procesos de la cuenca vertiente y del embalse y, que aprende de casos generados sintéticamente mediante un modelo hidrológico agregado y un modelo de optimización de la gestión del embalse. En una primera etapa, se generó un gran número de episodios sintéticos de avenida utilizando el método de Monte Carlo, para obtener las lluvias, y un modelo agregado compuesto de transformación lluvia- escorrentía, para obtener los hidrogramas de avenida. Posteriormente, se utilizaron las series obtenidas como señales de entrada al modelo de gestión de embalses PLEM, que optimiza una función objetivo de costes mediante programación lineal entera mixta, generando igual número de eventos óptimos de caudal vertido y de evolución de niveles en el embalse. Los episodios simulados fueron usados para entrenar y evaluar dos modelos de red Bayesiana, uno que pronostica el caudal de entrada al embalse, y otro que predice el caudal vertido, ambos en un horizonte de tiempo que va desde una a cinco horas, en intervalos de una hora. En el caso de la red Bayesiana hidrológica, el caudal de entrada que se elige es el promedio de la distribución de probabilidad de pronóstico. En el caso de la red Bayesiana hidráulica, debido al comportamiento marcadamente no lineal de este proceso y a que la red Bayesiana devuelve un rango de posibles valores de caudal vertido, se ha desarrollado una metodología para seleccionar un único valor, que facilite el trabajo del operador de la presa. Esta metodología consiste en probar diversas estrategias propuestas, que incluyen zonificaciones y alternativas de selección de un único valor de caudal vertido en cada zonificación, a un conjunto suficiente de episodios sintéticos. Los resultados de cada estrategia se compararon con el método MEV, seleccionándose las estrategias que mejoran los resultados del MEV, en cuanto al caudal máximo vertido y el nivel máximo alcanzado por el embalse, cualquiera de las cuales puede usarse por el operador de la presa en tiempo real para el embalse de estudio (Talave). La metodología propuesta podría aplicarse a cualquier embalse aislado y, de esta manera, obtener, para ese embalse particular, diversas estrategias que mejoran los resultados del MEV. Finalmente, a modo de ejemplo, se ha aplicado la metodología a una avenida sintética, obteniendo el caudal vertido y el nivel del embalse en cada intervalo de tiempo, y se ha aplicado el modelo MIGEL para obtener en cada instante la configuración de apertura de los órganos de desagüe que evacuarán el caudal. Currently, the dam operator for the management of dams uses simulation models during flood events, mainly due to its ease of use in real time. Some models have been developed to optimize the management of the reservoir to improve the results of simulation models. However, real-time application becomes very difficult or simply unworkable, because the decision to discharge depends on the unknown future avenue entering the reservoir. For this reason, the main goal is to develop a model of reservoir management at avenues that incorporates the advantages of an optimization model. At the same time, it should be easy to use in real-time by the dam manager. For this purpose, a Bayesian network model has been developed to represent the processes of the watershed and reservoir. This model learns from cases generated synthetically by a hydrological model and an optimization model for managing the reservoir. In a first stage, a large number of synthetic flood events was generated using the Monte Carlo method, for rain, and rain-added processing model composed of runoff for the flood hydrographs. Subsequently, the series obtained were used as input signals to the reservoir management model PLEM that optimizes a target cost function using mixed integer linear programming. As a result, many optimal discharge rate events and water levels in the reservoir levels were generated. The simulated events were used to train and test two models of Bayesian network. The first one predicts the flow into the reservoir, and the second predicts the discharge flow. They work in a time horizon ranging from one to five hours, in intervals of an hour. In the case of hydrological Bayesian network, the chosen inflow is the average of the probability distribution forecast. In the case of hydraulic Bayesian network the highly non-linear behavior of this process results on a range of possible values of discharge flow. A methodology to select a single value has been developed to facilitate the dam operator work. This methodology tests various strategies proposed. They include zoning and alternative selection of a single value in each discharge rate zoning from a sufficient set of synthetic episodes. The results of each strategy are compared with the MEV method. The strategies that improve the outcomes of MEV are selected and can be used by the dam operator in real time applied to the reservoir study case (Talave). The methodology could be applied to any single reservoir and, thus, obtain, for the particular reservoir, various strategies that improve results from MEV. Finally, the methodology has been applied to a synthetic flood, obtaining the discharge flow and the reservoir level in each time interval. The open configuration floodgates to evacuate the flow at each interval have been obtained applying the MIGEL model.

Analysis of the Optimal Sharing of Construction Risk in Public Procurement Contracts

This paper discusses a model based on the agency theory to analyze the optimal transfer of construction risk in public works contracts. The base assumption is that of a contract between a principal (public authority) and an agent (firm), where the payment mechanism is linear and contains an incentive mechanism to enhance the effort of the agent to reduce construction costs. A theoretical model is proposed starting from a cost function with a random component and assuming that both the public authority and the firm are risk averse. The main outcome of the paper is that the optimal transfer of construction risk will be lower when the variance of errors in cost forecast, the risk aversion of the firm and the marginal cost of public funds are larger, while the optimal transfer of construction risk will grow when the variance of errors in cost monitoring and the risk aversion of the public authority are larger

New methods for the synthesis of radiation patterns of coupled antenna arrays = Nuevos métodos de síntesis de diagramas de radiación de agrupaciones de antenas acopladas

El principal objetivo de esta tesis es el desarrollo de métodos de síntesis de diagramas de radiación de agrupaciones de antenas, en donde se realiza una caracterización electromagnética rigurosa de los elementos radiantes y de los acoplos mutuos existentes. Esta caracterización no se realiza habitualmente en la gran mayoría de métodos de síntesis encontrados en la literatura, debido fundamentalmente a dos razones. Por un lado, se considera que el diagrama de radiación de un array de antenas se puede aproximar con el factor de array que únicamente tiene en cuenta la posición de los elementos y las excitaciones aplicadas a los mismos. Sin embargo, como se mostrará en esta tesis, en múltiples ocasiones un riguroso análisis de los elementos radiantes y del acoplo mutuo entre ellos es importante ya que los resultados obtenidos pueden ser notablemente diferentes. Por otro lado, no es sencillo combinar un método de análisis electromagnético con un proceso de síntesis de diagramas de radiación. Los métodos de análisis de agrupaciones de antenas suelen ser costosos computacionalmente, ya que son estructuras grandes en términos de longitudes de onda. Generalmente, un diseño de un problema electromagnético suele comprender varios análisis de la estructura, dependiendo de las variaciones de las características, lo que hace este proceso muy costoso. Dos métodos se utilizan en esta tesis para el análisis de los arrays acoplados. Ambos están basados en el método de los elementos finitos, la descomposición de dominio y el análisis modal para analizar la estructura radiante y han sido desarrollados en el grupo de investigación donde se engloba esta tesis. El primero de ellos es una técnica de análisis de arrays finitos basado en la aproximación de array infinito. Su uso es indicado para arrays planos de grandes dimensiones con elementos equiespaciados. El segundo caracteriza el array y el acoplo mutuo entre elementos a partir de una expansión en modos esféricos del campo radiado por cada uno de los elementos. Este método calcula los acoplos entre los diferentes elementos del array usando las propiedades de traslación y rotación de los modos esféricos. Es capaz de analizar agrupaciones de elementos distribuidos de forma arbitraria. Ambas técnicas utilizan una formulación matricial que caracteriza de forma rigurosa el campo radiado por el array. Esto las hace muy apropiadas para su posterior uso en una herramienta de diseño, como los métodos de síntesis desarrollados en esta tesis. Los resultados obtenidos por estas técnicas de síntesis, que incluyen métodos rigurosos de análisis, son consecuentemente más precisos. La síntesis de arrays consiste en modificar uno o varios parámetros de las agrupaciones de antenas buscando unas determinadas especificaciones de las características de radiación. Los parámetros utilizados como variables de optimización pueden ser varios. Los más utilizados son las excitaciones aplicadas a los elementos, pero también es posible modificar otros parámetros de diseño como son las posiciones de los elementos o las rotaciones de estos. Los objetivos de las síntesis pueden ser dirigir el haz o haces en una determinada dirección o conformar el haz con formas arbitrarias. Además, es posible minimizar el nivel de los lóbulos secundarios o del rizado en las regiones deseadas, imponer nulos que evitan posibles interferencias o reducir el nivel de la componente contrapolar. El método para el análisis de arrays finitos basado en la aproximación de array infinito considera un array finito como un array infinito con un número finito de elementos excitados. Los elementos no excitados están físicamente presentes y pueden presentar tres diferentes terminaciones, corto-circuito, circuito abierto y adaptados. Cada una de estas terminaciones simulará mejor el entorno real en el que el array se encuentre. Este método de análisis se integra en la tesis con dos métodos diferentes de síntesis de diagramas de radiación. En el primero de ellos se presenta un método basado en programación lineal en donde es posible dirigir el haz o haces, en la dirección deseada, además de ejercer un control sobre los lóbulos secundarios o imponer nulos. Este método es muy eficiente y obtiene soluciones óptimas. El mismo método de análisis es también aplicado a un método de conformación de haz, en donde un problema originalmente no convexo (y de difícil solución) es transformado en un problema convexo imponiendo restricciones de simetría, resolviendo de este modo eficientemente un problema complejo. Con este método es posible diseñar diagramas de radiación con haces de forma arbitraria, ejerciendo un control en el rizado del lóbulo principal, así como en el nivel de los lóbulos secundarios. El método de análisis de arrays basado en la expansión en modos esféricos se integra en la tesis con tres técnicas de síntesis de diagramas de radiación. Se propone inicialmente una síntesis de conformación del haz basado en el método de la recuperación de fase resuelta de forma iterativa mediante métodos convexos, en donde relajando las restricciones del problema original se consiguen unas soluciones cercanas a las óptimas de manera eficiente. Dos métodos de síntesis se han propuesto, donde las variables de optimización son las posiciones y las rotaciones de los elementos respectivamente. Se define una función de coste basada en la intensidad de radiación, la cual es minimizada de forma iterativa con el método del gradiente. Ambos métodos reducen el nivel de los lóbulos secundarios minimizando una función de coste. El gradiente de la función de coste es obtenido en términos de la variable de optimización en cada método. Esta función de coste está formada por la expresión rigurosa de la intensidad de radiación y por una función de peso definida por el usuario para imponer prioridades sobre las diferentes regiones de radiación, si así se desea. Por último, se presenta un método en el cual, mediante técnicas de programación entera, se buscan las fases discretas que generan un diagrama de radiación lo más cercano posible al deseado. Con este método se obtienen diseños que minimizan el coste de fabricación. En cada uno de las diferentes técnicas propuestas en la tesis, se presentan resultados con elementos reales que muestran las capacidades y posibilidades que los métodos ofrecen. Se comparan los resultados con otros métodos disponibles en la literatura. Se muestra la importancia de tener en cuenta los diagramas de los elementos reales y los acoplos mutuos en el proceso de síntesis y se comparan los resultados obtenidos con herramientas de software comerciales. ABSTRACT The main objective of this thesis is the development of optimization methods for the radiation pattern synthesis of array antennas in which a rigorous electromagnetic characterization of the radiators and the mutual coupling between them is performed. The electromagnetic characterization is usually overlooked in most of the available synthesis methods in the literature, this is mainly due to two reasons. On the one hand, it is argued that the radiation pattern of an array is mainly influenced by the array factor and that the mutual coupling plays a minor role. As it is shown in this thesis, the mutual coupling and the rigorous characterization of the array antenna influences significantly in the array performance and its computation leads to differences in the results obtained. On the other hand, it is difficult to introduce an analysis procedure into a synthesis technique. The analysis of array antennas is generally expensive computationally as the structure to analyze is large in terms of wavelengths. A synthesis method requires to carry out a large number of analysis, this makes the synthesis problem very expensive computationally or intractable in some cases. Two methods have been used in this thesis for the analysis of coupled antenna arrays, both of them have been developed in the research group in which this thesis is involved. They are based on the finite element method (FEM), the domain decomposition and the modal analysis. The first one obtains a finite array characterization with the results obtained from the infinite array approach. It is specially indicated for the analysis of large arrays with equispaced elements. The second one characterizes the array elements and the mutual coupling between them with a spherical wave expansion of the radiated field by each element. The mutual coupling is computed using the properties of translation and rotation of spherical waves. This method is able to analyze arrays with elements placed on an arbitrary distribution. Both techniques provide a matrix formulation that makes them very suitable for being integrated in synthesis techniques, the results obtained from these synthesis methods will be very accurate. The array synthesis stands for the modification of one or several array parameters looking for some desired specifications of the radiation pattern. The array parameters used as optimization variables are usually the excitation weights applied to the array elements, but some other array characteristics can be used as well, such as the array elements positions or rotations. The desired specifications may be to steer the beam towards any specific direction or to generate shaped beams with arbitrary geometry. Further characteristics can be handled as well, such as minimize the side lobe level in some other radiating regions, to minimize the ripple of the shaped beam, to take control over the cross-polar component or to impose nulls on the radiation pattern to avoid possible interferences from specific directions. The analysis method based on the infinite array approach considers an infinite array with a finite number of excited elements. The infinite non-excited elements are physically present and may have three different terminations, short-circuit, open circuit and match terminated. Each of this terminations is a better simulation for the real environment of the array. This method is used in this thesis for the development of two synthesis methods. In the first one, a multi-objective radiation pattern synthesis is presented, in which it is possible to steer the beam or beams in desired directions, minimizing the side lobe level and with the possibility of imposing nulls in the radiation pattern. This method is very efficient and obtains optimal solutions as it is based on convex programming. The same analysis method is used in a shaped beam technique in which an originally non-convex problem is transformed into a convex one applying symmetry restrictions, thus solving a complex problem in an efficient way. This method allows the synthesis of shaped beam radiation patterns controlling the ripple in the mainlobe and the side lobe level. The analysis method based on the spherical wave expansion is applied for different synthesis techniques of the radiation pattern of coupled arrays. A shaped beam synthesis is presented, in which a convex formulation is proposed based on the phase retrieval method. In this technique, an originally non-convex problem is solved using a relaxation and solving a convex problems iteratively. Two methods are proposed based on the gradient method. A cost function is defined involving the radiation intensity of the coupled array and a weighting function that provides more degrees of freedom to the designer. The gradient of the cost function is computed with respect to the positions in one of them and the rotations of the elements in the second one. The elements are moved or rotated iteratively following the results of the gradient. A highly non-convex problem is solved very efficiently, obtaining very good results that are dependent on the starting point. Finally, an optimization method is presented where discrete digital phases are synthesized providing a radiation pattern as close as possible to the desired one. The problem is solved using linear integer programming procedures obtaining array designs that greatly reduce the fabrication costs. Results are provided for every method showing the capabilities that the above mentioned methods offer. The results obtained are compared with available methods in the literature. The importance of introducing a rigorous analysis into the synthesis method is emphasized and the results obtained are compared with a commercial software, showing good agreement.

A unified framework for linear function approximation of value functions in stochastic control

This paper contributes with a unified formulation that merges previ- ous analysis on the prediction of the performance ( value function ) of certain sequence of actions ( policy ) when an agent operates a Markov decision process with large state-space. When the states are represented by features and the value function is linearly approxi- mated, our analysis reveals a new relationship between two common cost functions used to obtain the optimal approximation. In addition, this analysis allows us to propose an efficient adaptive algorithm that provides an unbiased linear estimate. The performance of the pro- posed algorithm is illustrated by simulation, showing competitive results when compared with the state-of-the-art solutions.

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.