Examinando por Materia "ALGORITMOS GENÉTICOS"
Mostrando 1 - 20 de 20
Resultados por página
Opciones de ordenación
Ítem Compendium of M.Sc. publications on numerical estimation of effective properties of porous materials(Universidad EAFIT, 2014) Osorno Tejada, María Camila; Ruíz Salguero, Oscar EduardoÍtem Diseño de nuevos materiales duros por métodos AB Initio(Universidad EAFIT, 2013) Villegas Pulgarín, Edward Yesid; Vélez Ruiz, Mario ElkinÍtem Diseño de un sistema de pronóstico de la demanda de nitrógeno y oxígeno para la compañía Cryogas S.A. basado en redes neuronales(Universidad EAFIT, 2014) Rosas Pabón, Daniel Mauricio; Ramírez Echeverri, Sergio AugustoLa dinámica del mercado en los últimos años ha llevado a Cryogas a optimizar sus costos de producción y evitar agotados, maximizando el servicio al cliente y la rentabilidad de sus productos -- La realización este proyecto es con el objetivo de identificar la metodología propuesta genera un pronóstico acertado frente a los métodos y resultados utilizados actualmente y mejorar los costos productivos mediante una producción con una mejor eficiencia energética -- La investigación parte de la identificación de variables que afecta la demanda de los productos con mayor irregularidad, para cuantificar y generar entradas para los diferentes modelos de redes neuronales -- Se realizaron varios pronósticos con diferentes modelos cuyos resultados se comparan con datos de control para seleccionar el de mejor desempeño y mayor exactitud -- Los resultados obtenidos con el modelo de pronóstico de mejor desempeño se comparan con realidad y el método usado actualmente por Cryogas para identificar el mejor -- Los resultados obtenidos permiten una mejora en los procedimientos de Cryogas que impactaran positivamente la rentabilidad de la compañía, se tendrán parámetros de generación de pronósticos basados en criterios sólidos y se buscara eliminar la subjetividad en esta actividadÍtem Diseño y construcción de un sistema de levitación magnética controlado por un algoritmo PID(Universidad EAFIT, 2010) Ríos Ruiz, Juan David; Velásquez Torres, Álvaro AndrésÍtem Environmental improvement of operating supply chains: a multi-objective approach for the cement industry(Universidad EAFIT, 2017) Cadavid Giraldo, Nora; Vélez Gallego, Mario CésarNowadays companies worldwide face a growing pressure to reduce the environmental impact of their manufacturing activities -- However, the strategies used to achieve this goal are not clearly defined because of their conflicting relations with financial outcomes -- In parallel, globalization trends imply that as companies grow, usually through mergers and acquisitions, their supply chains become more complex -- The environmental improvement of these supply chains imply not only technical retrofit decisions aiming at adopting cleaner production technologies but also decisions regarding the structure of the supply chain itself -- Making these decisions becomes a difficult task because of the large number of variables involved, and the diversity of the interactions among them -- To tackle this problem, this research aims at providing a multi-objective solution approach for making technological retrofit decisions within an operating supply chain, so that both environmental and financial goals are best met -- The proposed solution approach is applied to the case of an operating cement supply chain in Colombia -- Several computational experiments were conducted, obtained results demonstrates that the proposed model is an e effective tool for multi-objective improvement decisions making, towards a more sustainable production processÍtem Estimación de parámetros de un UAV tipo ala volante mediante algoritmos metaheurísticos(Universidad EAFIT, 2014) Benjumea Giraldo, Jorge Alejandro; Vélez Sánchez, Carlos MarioEn este trabajo se describe el desarrollo y prueba de una herramienta en Matlab/Simulink para la estimación de parámetros aerodinámicos de un vehículo aéreo no tripulado (UAV) tipo ala volante a partir de las entradas y salidas observables del sistema, por medio de métodos metaheurísticos -- Para ello se procede con la descripción de un modelo matemático tipo caja gris implementado en Simulink, el cual representa la dinámica general de una aeronave descrita por las derivadas de estabilidad como coeficientes de las fuerzas y momentos inherentes al cuerpo rígido del sistema -- Igualmente se describen los algoritmos metaheurísticos utilizados por el software de estimación (algoritmos genéticos y optimización por enjambre de partículas), los cuales brindan las características más representativas del comportamiento social y evolutivo presente en esta disciplina -- A continuación, se realizan simulaciones con un modelo generador de datos con perturbaciones y un modelo caja gris para la estimación, de manera que se puedan validar la aplicabilidad, eficacia, problemas y bondades del método -- Finalmente, se realizan estimaciones de algunos parámetros en condiciones de vuelo longitudinal y se comparan los resultados con los datos teóricos -- Los resultados, obtenidos a partir de datos simulados con un modelo de un ala volante, muestran que la herramienta, principal aporte de este trabajo, permite estimar los parámetros de una manera amigable y eficiente, con desviaciones y tiempos de cómputo aceptables para once parámetrosÍtem FEA-driven Geometric Modelling for Meshless Methods(Springer Paris, 2005-11) Ruíz, Óscar; Granados, Miguel; Cadavid, Carlos; Universidad EAFIT. Departamento de Ingeniería Mecánica; Laboratorio CAD/CAM/CAEOptimized Boolean Operations against orthogonal Fixed Grids (FG) for 2-manifold construction in quasi-meshless methods for Finite Element Analysis are presented -- A Piecewise Linear (PL) or Boundary Representation (B-Rep) B is assumed to be the boundary of a solid S ⊂ R3 -- On the other hand, R3 is partitioned into a 3-dimensional array of cubic, uniform cells Ci,j,k . Cells Ci,j,k with Ci,j,k ∩ S ≠Φ and Ci,j,k ∩ S ≠ Ci,j,k are particularly important for FG applications -- These are the cells Ci,j,k intersecting B, which happen to be Neither Inside nor Outside (NIO) of B -- The boundary ∂(Ci,j,k ∩ S ) of Ci,j,k ∩ S must be calculated from ∂Ci,j,k and B for a large number of cells Ci,j,k , which makes the normal boolean operations unpractical -- The article illustrates with examples the immersion of B-Rep models in Fixed Grids, visits the downstream results of the stress-strain calculations using FG and explains how this approach is used in Product Design OptimizationÍtem Genetic algorithm with a Bayesian approach for multiple change-point detection in time series of counting exceedances for specific thresholds(Universidad EAFIT, 2023) Taimal Yepes, Carlos Alberto; Suárez Sierra, Biviana MarcelaAlthough the applications of Non-Homogeneous Poisson Processes (NHPP) to model and study the threshold overshoots of interest in different time series of measurements have proven to provide good results, they needed to be complemented with an eficient and automatic diagnostic technique to establish the location of the change-points, which, when taken into account, make the estimated model it poorly in regards of the information contained in the real one. Because of this, a new method is proposed to solve the segmentation uncertainty of the time series of measurements, where the generating distribution of exceedances of a specific threshold is the focus of investigation. One of the great contributions of the present algorithm is that all the days that trespassed are candidates to be a change-point, so all the possible configurations of overflow days under the heuristics of a genetic algorithm are the possible chromosomes, which will unite to produce new solutions. Also, such methods will be guarantee to non-local and the best possible one solution, reducing wasted machine time evaluating the least likely chromosomes to be a feasible solution. The analytical evaluation technique will be by means of the Minimum Description Length (MDL) as the objective function, which is the joint posterior distribution function of the parameters of the NHPP of each regime and the change-points that determines them and which account as well for the influence of the presence of said times. Thus, one of the practical implications of the present work comes in terms of overcoming the need of modeling the time series of measurements, where the distributions of exceedances of certain thresholds, or where the counting of certain events involving abrupt changes, is the main focus with applications in phenomena such as climate change, information security and epidemiology, to name a few.Ítem Heuristic to allocate intermediate buffer storage capacities in a production line subject to machine breakdowns(Production and Operations Management Society (POMS), 2013-05-03) Vélez Gallego, Mario César; Jaramillo Jiménez, Jhull Breynner; Universidad EAFIT. Departamento de Ingeniería de Producción; marvelez@eafit.edu.co; Gestión de Producción y LogísticaIn this research proposal we consider a production line subject to random failures at each workstation and operating under a make-to-stock policy -- Every time a workstation fails, a corrective maintenance activity is triggered to repair the workstation -- In order to palliate the effect of the random failures in the performance of the system, intermediate buffers are placed in-between workstations -- An inventory holding cost is associated to eachbuffer -- The research objective in this work is to allocate capacity to each intermediate buffer in the line so that the average cost per time unit is minimized while the average service level is kept above a minimum pre-specified value -- In this paper we assume that unsatisfied demand is lost and the service level is defined as the long term proportion of satisfied demand -- A greedy simulation–based heuristic is presented to find a feasible solution to the problemÍtem Makespan minimization in a job shop with a BPM using simulated annealing(Springer London, 2013-03-08) Rojas Santiago, Miguel; Vélez Gallego, Mario César; Damodaran, Purushothaman; Muthuswamy, Shanthi; Rojas Santiago, Miguel; Vélez Gallego, Mario César; Damodaran, Purushothaman; Muthuswamy, Shanthi; Universidad EAFIT. Departamento de Ingeniería de Producción; marvelez@eafit.edu.co; Gestión de Producción y LogísticaA scheduling problem commonly observed in the metal working industry has been studied in this research effort -- A job shop equipped with one batch processing machine (BPM) and several unit-capacity machines has been considered. Given a set of jobs, their process routes,processing requirements, and size, the objective is to schedule the jobs such that the makespan is minimized -- The BPM can process a batch of jobs as long as its capacity is not exceeded -- The batch processing time is equal to the longest processing job in the batch -- If no batches were to be formed, the scheduling problem under study reduces to the classicaljob shop problem with makespan objective, which is known to be nondeterministic polynomial time-hard -- A network representation of the problem using disjunctive and conjunctive arcs, and a simulated annealing (SA) algorithm are proposed to solve the problem. The solution quality and run time of SA are compared with CPLEX, a commercial solver used to solve the mathematical formulation and with four dispatching rules -- Experimental study clearly highlights the advantages, in terms of solution quality and run time, of using SA to solve large-scale problemsÍtem Metaheurísticos: una alternativa para la solución de problemas combinatorios en Administración de Operaciones(Revista EIA, 2007-12) Vélez, Mario César; Montoya, José Alejandro; Universidad EAFIT. Departamento de Ingeniería de Producción; marvelez@ eafit.edu.co; jmonto36@eafit.edu.co; Gestión de Producción y LogísticaThe scarce diffusion given to the newest techniques for solving complex operations management problems has as a direct consequence that companies lose opportunities to operate at lower costs and higher efficiency -- The objective of this article is to introduce and explain the fundamental ideas behind metaheuristics, a solution technique for combinatorial problems that has received the most attention from the academiccommunity in the last few years -- In order to illustrate these ideas, an example of a classical combinatorial problem in the sequencing of operations area is presented, and a solution algorithm making use of some of these techniques is proposedÍtem Métodos computacionales para la solución de problemas de optimización bajo incertidumbre y problemas inversos, desde la perspectiva intervalo-valuada(Universidad EAFIT, 2011) Cano Cadavid, Andrés Felipe; Puerta Yepes, María EugeniaÍtem La minería de datos y su evolución a la GRID(Universidad EAFIT, 2010) Serna Tangarife, Andrés Santiago; Cardona Ríos, SoniaÍtem Optimización de empaques aplicando lineamientos de diseño para el ensamble validado en los contextos de empaque, transporte y desempaque(Universidad EAFIT, 2013) Betancur Muñoz, Pamela; Martínez Cadavid, José Fernando; Duque Lombana, Juan FernandoÍtem Optimización de muros de contención en voladizo, mediante algoritmos genéticos(Universidad EAFIT, 2011) Sayago Alzate, Andrés Felipe; Múnera Guerra, Luis Miguel; Guerra Londoño, Jorge EduardoÍtem Programación lineal y algoritmos genéticos para la solución de un problema de corte(Universidad EAFIT, 2008) Jaramillo Jaramillo, Juan David; Correa Zabala, Francisco José; Jaramillo Mejía, RicardoEste proyecto de grado discute el problema de corte (roll-trim o cutting stock) en el que se busca optimizar la cantidad de material utilizada en un proceso de producción -- Por la naturaleza del problema, el enfoque tradicional de la programación lineal no es muy efectivo -- Una buena solución al problema debe considerar el desperdicio de material, los cambios de patrones de corte en la máquina y la cantidad de material procesado -- Proponemos una solución utilizando un algoritmo genético que tiene en cuenta las consideraciones anteriores y probamos que su desempeño es superior a la solución obtenida por el enfoque como problema de programación linealÍtem Propiedades estructurales y electrónicas en cristales de MoC basados en DFT(Universidad EAFIT, 2016) Pérez Rave, Hernando Enrique; David Caro, Jorge LeónÍtem Solar beam radiation modeling for design and simulation of V-Trough photovoltaic applications(Universidad EAFIT, 2017) Arias Rosales, Andrés; Mejía-Gutiérrez, RicardoPhotovoltaic V-Troughs use simple and low-cost non-imaging optics, namely flat mirrors, to increase the solar harvesting area by concentrating the solar rays towards the photovoltaic surface -- Combined with stepped manual tracking, these devices can boost the performance of regular solar cells for residential and personalized scenarios in developing countries -- The geometrical set-up of a V-Trough, related to its length and angular parameters, conditions the optical phenomena that occur as the solar beam radiation dynamically interacts with the surfaces of the device -- These phenomena, such as shadows and consecutive reflections, determine the solar radiation that is effectively concentrated towards the photovoltaic area -- To support the design and simulation of such devices, this research project proposes a theoretical model of their effective concentration, seeking a high flexibility in the geometrical inputs, a detailed output of the optical phenomena involved and a low computational demand -- The model is based on the geometrical optics of beam solar radiation and it was experimentally validated in a statistical comparison against experimental measurements performed in a laser testing platform -- Based on the model, several design performance indicators were established: i) Ce (the average effective concentration); ii) Cost (the cost of materials); iii) Tsp (the space required) -- Moreover, from the integration of these indicators in proportional comparisons, three multi-objective indices were proposed: a) ICOE (Ce versus Cost); b) MICOE (Ce versus Cost and Ce versus Tsp combined); c) MDICOE (the same comparison as the previous one but also including the discretization effects related to the solar cells) – As tools for supporting the V-Trough set-up definition process, an interactive software and a series of genetic algorithms were developed based on these indicators and indices and the modeling framework developed -- The proposed software, named “VTDesign”, allows a V-Trough designer to iterate and compare diverse design alternatives in real-time by means of a visualization of the optical performance, representations of the geometrical set-ups and a cost-effectiveness analysis -- Among the genetic algorithms studied, a new genetic algorithm, named “GA-WA” (Genetic Algorithm - Weibull Arias), was proposed -- This tool uses heuristic processes, based on biomimicry and Weibull probability distributions, in order to optimize the indicators and indices as fitness/objective functions -- The intuition and the more holistic exploration of a trained engineer with VTDesign can be complemented with the broader and less biased evolutionary optimization of GAWA -- In a design case study, both VTDesign and GA-WA were implemented and found to be effective in the indices improvement, as well as efficient and flexible tools in the problem of defining the set-up of a solar V-Trough in a given personalized scenarioÍtem Structural characterization of the (MeSH) 4 potential energy surface(Springer-Verlag, 2013-01-10) Guerra Tamayo, Doris Lucía; Restrepo Cossio, Albeiro Alonso; Gómez Maya, Sara Luz; David Caro, Jorge León; Universidad EAFIT. Departamento de Ciencias Básicas; Jorge León David Caro (jdavidca@eafit.edu.co); Electromagnetismo Aplicado (Gema)A random walk on the PES for (MeSH)4 clusters produced 50 structural isomers held together by hydrogenbonding networks according to calculations performed at the B3LYP/6–311++G** and MP2/6–311++G** levels -- The geometric motifs observed are somewhat similar to those encountered for the methanol tetramer, but the interactions responsible for cluster stabilization are quite different in origin -- Cluster stabilization is not related to the number of hydrogen bonds -- Two distinct, well-defined types of hydrogen bonds scattered over a wide range of distances are predictedÍtem The classification problem in machine learning: an overview with study cases in emotion recognition and music-speech differentiation(Universidad EAFIT, 2015) Rodríguez Cadavid, SantiagoThis work addresses the well-known classification problem in machine learning -- The goal of this study is to approach the reader to the methodological aspects of the feature extraction, feature selection and classifier performance through simple and understandable theoretical aspects and two study cases -- Finally, a very good classification performance was obtained for the emotion recognition from speech