No description available.

Le problème de localisation-routage avec capacités (PLRC) apparaît comme un problème clé dans la conception de réseaux de distribution de marchandises. Il généralisele problème de localisation avec capacités (PLC) ainsi que le problème de tournées de véhicules à multiples dépôts (PTVMD), le premier en ajoutant des décisions liées au routage et le deuxième en ajoutant des décisions liées à la localisation des dépôts. Dans cette thèse on dévelope des outils pour résoudre le PLRC à l’aide de la programmation mathématique. Dans le chapitre 3, on introduit trois nouveaux modèles pour le PLRC basés sur des flots de véhicules et des flots de commodités, et on montre comment ceux-ci dominent, en termes de la qualité de la borne inférieure, la formulation originale à deux indices [19]. Des nouvelles inégalités valides ont été dévelopées et ajoutées aux modèles, de même que des inégalités connues. De nouveaux algorithmes de séparation ont aussi été dévelopés qui dans la plupart de cas généralisent ceux trouvés dans la litterature. Les résultats numériques montrent que ces modèles de flot sont en fait utiles pour résoudre des instances de petite à moyenne taille. Dans le chapitre 4, on présente une nouvelle méthode de génération de colonnes basée sur une formulation de partition d’ensemble. Le sous-problème consiste en un problème de plus court chemin avec capacités (PCCC). En particulier, on utilise une relaxation de ce problème dans laquelle il est possible de produire des routes avec des cycles de longueur trois ou plus. Ceci est complété par des nouvelles coupes qui permettent de réduire encore davantage le saut d’intégralité en même temps que de défavoriser l’apparition de cycles dans les routes. Ces résultats suggèrent que cette méthode fournit la meilleure méthode exacte pour le PLRC. Dans le chapitre 5, on introduit une nouvelle méthode heuristique pour le PLRC. Premièrement, on démarre une méthode randomisée de type GRASP pour trouver un premier ensemble de solutions de bonne qualité. Les solutions de cet ensemble sont alors combinées de façon à les améliorer. Finalement, on démarre une méthode de type détruir et réparer basée sur la résolution d’un nouveau modèle de localisation et réaffectation qui généralise le problème de réaffectaction [48]. / The capacitated location-routing problem (CLRP) arises as a key problem in the design of distribution networks. It generalizes both the capacitated facility location problem (CFLP) and the multiple depot vehicle routing problem (MDVRP), the first by considering additional routing decisions and the second by adding the location decision variables. In this thesis we use different mathematical programming tools to develop and specialize new models and algorithms for solving the CLRP. In Chapter 3, three new models are presented for the CLRP based on vehicle-flow and commodity-flow formulations, all of which are shown to dominate, in terms of the linear relaxation lower bound, the original two-index vehicle-flow formulation [19]. Known valid inequalities are complemented with some new ones and included using separation algorithms that in many cases generalize extisting ones found in the literature. Computational experiments suggest that flow models can be efficient for dealing with small or medium size instances of the CLRP (50 customers or less). In Chapter 4, a new branch-and-cut-and-price exact algorithm is introduced for the CLRP based on a set-partitioning formulation. The pricing problem is a shortest path problem with resource constraints (SPPRC). In particular, we consider a relaxation of such problem in which routes are allowed to contain cycles of length three or more. This is complemented with the development of new valid inequalities that are shown to be effective for closing the optimality gap as well as to restrict the appearance of cycles. Computational experience supports the fact that this method is now the best exact method for the CLRP. In Chapter 5, we introduce a new metaheuristic with the aim of finding good quality solutions in short or moderate computing times. First, a bundle of good solutions is generated with the help of a greedy randomized adaptive search procedure (GRASP). Following this, a blending procedure is applied with the aim of producing a better upper bound as a combination of all the others in the bundle. An iterative destroy-and-repair method is then applied using a location-reallocation model that generalizes the reallocation model due to de Franceschi et al. [48].

La traduction automatique statistique est un domaine très en demande et où les machines sont encore loin de produire des résultats de qualité humaine. La principale méthode utilisée est une traduction linéaire segment par segment d'une phrase, ce qui empêche de changer des parties de la phrase déjà traduites. La recherche pour ce mémoire se base sur l'approche utilisée dans Langlais, Patry et Gotti 2007, qui tente de corriger une traduction complétée en modifiant des segments suivant une fonction à optimiser. Dans un premier temps, l'exploration de nouveaux traits comme un modèle de langue inverse et un modèle de collocation amène une nouvelle dimension à la fonction à optimiser. Dans un second temps, l'utilisation de différentes métaheuristiques, comme les algorithmes gloutons et gloutons randomisés permet l'exploration plus en profondeur de l'espace de recherche et permet une plus grande amélioration de la fonction objectif. / Statistical Machine Translation is a field ingreat demand and where machines are still far from producing human-level results.The main method used is a segment by segment linear translation of a sentence, which prevents modification of already translated parts of the sentence. Research for this memoir is based on an approach used by Langlais, Patry and Gotti 2007, which tries to correct a completed translation by modifying segments following a function which needs to be optimized. As a first step, exploration of new traits such as an inverted language model and a collocation model brings a new dimension to the optimization function. As a second step, use of different metaheuristics, such as the greedy and randomized greedy algorithms, allows greater depth while exploring the search space and allows a greater improvement of the objective function.

La dispersion spatiale de l'habitat et des activités de ces dernières décennies a fortement contribué à un allongement des distances et des temps de trajets domicile-travail. Cela a pour conséquence un accroissement de l'utilisation des voitures particulières, notamment au sein et aux abords des grandes agglomérations. Afin de réduire les impacts dus à l'augmentation du trafic routier, des services de covoiturage, où des usagers ayant la même destination se regroupent en équipage pour se déplacer, ont été mis en place partout dans le monde. Nous présentons ici nos travaux sur le problème de covoiturage régulier. Dans cette thèse, le problème de covoiturage régulier a été modélisé et plusieurs métaheuristiques de résolution ont été implémentées, testées et comparées. La thèse est organisée de la façon suivante: tout d'abord, nous commençons par présenter la définition et la description du problème ainsi que le modèle mathématique associé. Ensuite, plusieurs métaheuristiques pour résoudre le problème sont présentées. Ces approches sont au nombre de quatre: un algorithme de recherche locale à voisinage variable, un algorithme à base de colonies de fourmis, un algorithme génétique guidée et un système multi-agents génétiques auto-adaptatif. Des expériences ont été menées pour démontrer l'efficacité de nos approches. Nous continuons ensuite avec la présentation et la résolution d'une extension du problème de covoiturage occasionel comportant plusieurs destinations. Pour terminer, une plate-forme de test et d'analyse pour évaluer nos approches et une plate-forme de covoiturage sont présentées dans l'annexe. / Nowadays, the increased human mobility combined with high use of private cars increases the load on environment and raises issues about quality of life. The extensive use of private cars lends to high levels of air pollution, parking problem, traffic congestion and low transfer velocity. In order to ease these shortcomings, the car pooling program, where sets of car owners having the same travel destination share their vehicles, has emerged all around the world. We present here our research on the long-term car pooling problem. In this thesis, the long-term car pooling problem is modeled and metaheuristics for solving the problem are investigated. The thesis is organized as follows. First, the definition and description of the problem as well as its mathematical model are introduced. Then, several metaheuristics to effectively and efficiently solve the problem are presented. These approaches include a Variable Neighborhood Search Algorithm, a Clustering Ant Colony Algorithm, a Guided Genetic Algorithm and a Multi-agent Self-adaptive Genetic Algorithm. Experiments have been conducted to demonstrate the effectiveness of these approaches on solving the long-term car pooling problem. Afterwards, we extend our research to a multi-destination daily car pooling problem, which is introduced in detail manner along with its resolution method. At last, an algorithm test and analysis platform for evaluating the algorithms and a car pooling platform are presented in the appendix.

O objetivo deste trabalho é apresentar uma nova abordagem para obtenção de configurações para sistemas de distribuição de energia elétrica com o intuito de minimizar o valor de perdas ativas sem violar as restrições operacionais. Para isso, considera-se que os sistemas de distribuição estão operando em regime permanente e que suas fases estão equilibradas e simétricas, podendo o sistema ser representado por um diagrama unifilar. A reconfiguração é feita de forma a redistribuir os fluxos de corrente nas linhas, transferindo cargas entre os alimentadores e melhorando o perfil de tensão ao longo do sistema. O problema de reconfiguração do sistema pode ser formulado como um problema de programação não-linear inteiro misto. Devido à explosão combinatorial inerente a este tipo de problema, a resolução do mesmo por técnicas de otimização clássicas torna-se pouco atraente, dando espaço para técnicas heurísticas e metaheurísticas. Essas outras, mesmo não garantindo o ótimo global, são capazes de encontrar boas soluções em um espaço de tempo relativamente curto. Para a resolução do problema de reconfiguração, utilizou-se uma nova metodologia baseada no comportamento de colônias de formigas em busca de alimento na natureza. Nesta, formigas artificiais (agentes) exploram o meio ambiente (sistema de distribuição) e trocam informações para tentar encontrar a topologia que apresente os menores valores de perdas ativas. Para o cálculo das perdas, este trabalho também apresenta uma nova abordagem para resolução do problema de fluxo de potência (FP) em sistemas de distribuição radial. O fluxo de potência é uma ferramenta básica utilizada pelos centros de controle para determinar os estados e condições operacionais desses sistemas de potência. Basicamente, as metodologias empregadas para o cálculo do fluxo de potência são baseadas nos métodos clássicos de Newton ou Gauss. Mas em sistemas de distribuição de energia, devido a particularidades inerentes a estes, como a alta relação entre resistência e reatância das linhas (r/x) e a operação radial, estes métodos apresentam problemas de convergência e se tornam ineficientes na maioria das vezes. A abordagem consiste na associação dos métodos da função penalidade e de Newton. O mal-condicionamento da matriz Jacobiana de Newton é resolvido pela associação com o método da função penalidade. São apresentados testes realizados em sistemas de 5 barras, 16 barras, 33 barras, 69 barras e 136 barras para avaliar a potencialidade das técnicas propostas. Os resultados são considerados bons ou muito bons quando comparado com as técnicas existentes atualmente. / The objective of this work is to present a novel methodology for obtaining new configurations of the distribution system in order to minimize the active power losses without violating operational constraints. For this, it is considered that any distribution system is operating in a steady state and that it is balanced, therefore it can be represented by a one-line diagram. The reconfiguration is done in order to redistribute de current flows on the distribution power lines, transferring loads among the feeders and improving the voltage profile along the system. Such problem can be formulated as a mixed integer nonlinear programming problem. Due to its inherent combinatorial characteristic and since its solution by classic optimization techniques is not appealing, heuristic and metaheuristic techniques are thus better suited for its solution. Although these latter do not guarantee a global optimum, they are able to find good solutions in a relatively short time. The solution of the reconfiguration problem in this approach makes use of a novel methodology based on ant colony behavior, when these search for victuals in nature. In this technique, the artificial ants (agents) explore the environment (distribution system) and exchange information among them in order to find the topology that provides the smallest active losses. For the active losses calculation, this work also presents a novel approach for the solution of the power flow problem for radial distribution systems. The solution of the power flow problem is used by system operators in order to determine the state and operational conditions of power systems. Basically, the most common techniques used in the power flow solution are based on either Newton\'s or Gauss\' approaches. However, due to particular characteristics of distribution systems such as the high ratio of r/x and the radial topology, these methods present convergence problems and are not efficient in most of the cases. Thus, this novel technique consists in associating Newton\'s and the penalty function approaches. The matter of the ill-conditioned Jacobian matrix in Newton\'s method is overcome with the penalty function method. Some tests performed in different systems are then presented in order to assess the effectiveness of both proposed techniques.

Neste trabalho é proposta uma nova técnica de solução para resolver o problema de planejamento da expansão de sistemas de transmissão estático através da introdução da tomada de decisão fuzzy. Na técnica apresentada neste trabalho, a tomada de decisão fuzzy é aplicada para o desenvolvimento de um algoritmo heurístico construtivo. O sistema fuzzy é utilizado para contornar alguns problemas críticos das heurísticas que utilizam o índice de sensibilidade como guia para inserção de novas linhas. A heurística apresentada nesse trabalho é baseada na técnica dividir para conquistar. Verificou-se que a deficiência das heurísticas construtivas é decorrente da decisão de inserir novas linhas baseada em valores não seguros encontrados através da solução do modelo utilizado. Para contornar tal deficiência, sempre que surgirem valores não seguros divide-se o problema original em dois subproblemas, um que analisa a qualidade da resposta para o caso em que a linha é inserida e outro para verificar a qualidade da resposta para o caso em que a linha não é inserida. A tomada de decisão fuzzy é utilizada para decidir sobre quando dividir o problema em dois novos subproblemas. Utilizou-se o modelo cc com a estratégia de Villasana-Garver-Salon para realizar a modelagem da rede elétrica para os problemas da expansão de sistemas de transmissão aqui propostos. Ao serem realizados testes em sistemas de pequeno, médio e grande portes certificou-se que o método pode encontrar a solução ótima de sistemas de pequeno e médio portes. Porém, a solução ótima dos sistemas de grande porte testados não foi encontrada. Para melhorar a qualidade da solução encontrada utilizou, em uma segunda fase, a metaheurística busca tabu. A busca tabu utiliza o modelo cc. Os resultados se mostraram bastante promissores. Os testes foram realizados em alguns sistemas reais brasileiros e com o sistema real colombiano. / A new solution technique to solve the long-term static transmission expansion planning (TEP) problem based on fuzzy decision making is proposed. The technique applies the concepts of fuzzy decision making in a constructive heuristic algorithm. The fuzzy system is used to circumvent some critical problems of heuristics that use sentivity indices as a guide for insertion and construction of new lines. The heuristic algorithm proposed in this work is based on the divide and conquer technique. It has been verified that the deficiency of the constructive heuristics is due to the decision of inserting new lines based only on information given by the index, which usually is calculated from a relaxed mathematical representation of the problem and can become less accurate during the solution process. In order to be able to deal with such problem, whenever the quality of the index decreases, the original problem is divided into two sub-problems: one examines the quality of the solution when the transmission line indicated by the sensitivity index is inserted and the other subproblem checks the opposite. Fuzzy decision-making is used to decide the moment to divide the problem into two subproblems based on other information. The hybrid linear model is used to model the long-term transmission expansion planning problem and is used in the proposed algorithm. Tests was done with systems of small-term, medium-term and long-term. The optimal solution of small-term and medium-term was foundo using just the construtive heuristic algorithm with fuzzy decision-making. To deal with long-term systems was used the solutions of the construtive heuristic algorithm with fuzzy decision-making to init a tabu search. The tabu search uses the dc model. The results are very promising. The test was done with some real brazilian systems and with the real colombian system.

Dans les problèmes de localisation-routage classiques (LRP), il s'agit de combiner des décisions stratégiques liées aux choix des sites à ouvrir (centres de traitement) avec des décisions tactiques et opérationnelles liées à l'affectation des clients aux sites sélectionnés et a la confection des tournées associées. Cette thèse propose de nouveaux modèles de localisation-routage permettant de résoudre des problématiques issues de réseaux logistiques, devenus aujourd'hui de plus en plus complexes vu la nécessité de mutualisation de ressources pour intégrer des contraintes de développement durable et des prix de carburants qui semblent augmenter de manière irrémédiable. Plus précisément, trois aspects ont été intégrés pour généraliser les modèles LRP classiques de la littérature : 1) l'aspect pickup and delivery, 2) l'aspect multi-produits, et 3) la possibilité de visiter un ou plusieurs centres de traitement dans une tournée donnée. Nous avons étudié deux schémas logistiques, qui ont donné lieu à deux nouveaux modèles de localisation et de routage, le MPLRP-PD (LRP with multi-product and pickup and delivery), qui peut être vu comme une extension des problèmes de tournées de véhicules avec collecte et livraison, intégrant une décision tactique liée à la localisation des centres de traitement (noeud avec collecte et livraison) dans un réseau de distribution à un seul échelon, et le 2E-MPLRP-PD (Two-echelon LRP with multi-product and pickup and delivery) qui est une généralisation du LRP à deux échelons avec les contraintes citées plus-haut. Ces deux modèles ont été formalisés par des programmes linéaires en variables mixtes (MIP). Des techniques de résolution, basées sur des méthodes de type heuristique, clustering, métaheuristique, ont été proposées pour résoudre le MPLRP-PD et le 2E-MPLRP-PD. Les jeux d'essais de la littérature ont été généralisés pour tester et valider les algorithmes proposés / In the framework of Location-Routing Problem (LRP), the main idea is to combine strategic decisions related to the choice of processing centers with tactical and operational decisions related to the allocation of customers to selected processing centers and computing the associated routes. This thesis proposes a new location-routing model to solve problems which are coming from logistics networks, that became nowadays increasingly complex due to the need of resources sharing, in order to integrate the constraints of sustainable development and fuels price, which is increasing irreversibly. More precisely, three aspects have been integrated to generalize the classical LRP models already existed in the literature: 1) pickup and delivery aspect, 2) multi-product aspect, and 3) the possibility to use the processing centers as intermediate facilities in routes. We studied two logistics schemes gives us two new location-routing models: (i) MPLRP-PD (Multi-product LRP with pickup and delivery), which can be viewed as an extension of the vehicle routing problem with pick-up and delivery, including a tactical decision related to the location of processing centers (node with pick-up and delivery), and (ii) 2E-MPLRP-PD (Two-echelon multi-product LRP with pickup and delivery), which is a generalization of the two-echelon LRP. Both models were formalized by mixed integer linear programming (MIP). Solving techniques, based on heuristic methods, clustering approach and meta-heuristic techniques have been proposed to solve the MPLRP-PD and the 2E-MPLRP-PD. The benchmarks from the literature were generalized to test and to validate the proposed algorithms

Le volet applicatif de cette thèse porte sur l'agencement d'un logement destiné à une personne en situation de handicap. L'agencement désigne le choix de la position, de la forme et des dimensions des pièces, des portes et des couloirs. L'agencement est généralement élaboré par un architecte, dans le respect d'un nombre si élevé de contraintes qu'il lui est difficile de parvenir qu'il parvienne à toutes les satisfaire : il y a d'abord des contraintes architecturales évidentes : non recouvrement des pièces, largeur suffisante des couloirs, accessibilité à tout point du lieu à partir de tout autre point, nécessité de placer certaines pièces sur des arrivées ou évacuations … Il y a ensuite les contraintes imposées par le handicap : largeur accrue des couloirs (déplacement en fauteuil), nécessité d'assurer un effort quotidien minimum (lutte contre le vieillissement), limitation des escaliers (asthme sévère), éloignement d'une pièce des murs mitoyens (surdité) .... Et il y a finalement les souhaits exprimés par le futur occupant, par exemple minimiser certains trajets, maximiser l’éloignement entre deux pièces ou imposer l’orientation d’une pièce. D’un point de vue formel, notre travail a consisté à développer d'une part des modèles mathématiques et des méthodes algorithmiques capables de gérer ces contraintes et d'autre part des prototypes logiciels opérationnels. Les méthodes élaborées relèvent de deux approches : l'optimisation d'un agencement conçu par l'architecte et la synthèse d'un agencement sans suggestion initiale de l'architecte. La synthèse d'un plan a été abordée comme un problème de type « bin-packing » (réputé NP-difficile) avec des contraintes additionnelles : les objets à placer - les pièces - ont des tailles variables et ils sont soumis à des contraintes fonctionnelles. La méthode de résolution s'appuie sur un premier modèle mathématique, qui prend la forme d’un programme quadratique (linéarisé par la suite) en variables mixtes. Elle a été appliquée avec succès pour placer les pièces d'un logement, pour les dimensionner, pour déterminer les couloirs assurant une complète accessibilité au logement et pour prendre en compte certaines contraintes imposées par le handicap du futur occupant. Un deuxième modèle mathématique a été élaboré pour le placement des portes et une heuristique a enfin été développée pour affecter l'espace occupé par les couloirs non indispensables aux pièces avoisinantes. La totalité de cette démarche a été programmée dans un prototype logiciel pleinement opérationnel. Le deuxième ensemble de contributions concerne l'optimisation d'un agencement existant. Cette optimisation a été conçue comme un processus itératif enchaînant évaluation et modification (amélioration) d'un agencement. Il est décliné de quatre manières : une métaheuristique de type « recuit simulé » et trois méthodes de type « recherche locale », qui explorent l’espace des solutions en utilisant des voisinages spécialement définis. Cette approche a d'une part permis d’appréhender le caractère multicritère de cette problématique et a d'autre part exigé la mise en œuvre de nombreux algorithmes géométriques. Ces travaux sont implantés dans un deuxième prototype logiciel. Ce projet a nécessité la participation à de nombreuses manifestations au-delà du domaine de l’informatique, nationales et régionales, scientifiques et non-scientifiques, organisées par différents organismes politiques et associatifs travaillant sur la problématique du handicap et de l’accessibilité, afin de bien appréhender les attentes du monde scientifique et socioprofessionnel. Cette phase prospective a été concrétisée par la rédaction de nombreux rapports qui ont alimentés la bibliographie du mémoire de thèse / At an application level, this thesis deals with the layout of an accommodation intended for a disabled person. Determining the layout means choosing the position, shape and dimensions of rooms, doors and corridors. It is usually an architect's job but the complexity is such that it is very unlikely that he succeeds in optimally fulfilling all the constraints: first, there are architectural constraints: no room overlapping, sufficient width for the corridors, accessibility to and from any point, mandatory positioning of some rooms on some areas (e.g. water supply and outlet) … Then, there are constraints imposed by disabilities: enlarged corridors (wheelchairs), mandatory daily amount of efforts (fight against aging), reducing the number of steps (severe asthma), moving a room away from shared walls (deafness)... Finally, there are the wishes expressed by the future occupant, such as minimizing some journeys, maximizing the distance between two rooms or fixing a room's orientation. From a formal point of view, our work has consisted, firstly, in developing mathematical models and algorithmic methods to deal with all these constraints and, secondly, in realizing software prototypes applying these concepts. The tools we propose aim either at optimizing a layout previously designed by an architect or at synthesising a layout without any initial suggestions from the architect. Synthesis has been tackled as bin-packing-type problem (known to be NP-hard) but with additional constraints: the objects to be placed (the rooms) have variable sizes and they are submitted to functional constraints. The resolution is based on a first, initially quadratic and then linearized, mixed integer mathematical model. It has been successfully applied to position and dimension the rooms of an accommodation, to determine corridors allowing a full accessibility to all the rooms and to take into account a number of constraints coming from the disabilities of the future occupant. A second mathematical model has been formulated for the positioning of the doors and, finally, a heuristic method has been designed to assign the space used by useless corridors to adjacent rooms. The whole process has been embedded in a fully operational software. The second set of contributions is about the optimization of an existing layout. This task has been tackled through an iterative process, looping on evaluation and modification (improvement) of an accommodation. It has been implemented in four different ways: a metaheuristic (simulated annealing) and three local-search-type methods, which traverse the solution space by using specific definitions of the neighbourhood. This approach has firstly underlined the multicriteria feature of our problem and, secondly, has required the development of many computational geometry algorithms. All this work is integrated in another functional prototype software. To understand the expectations of the scientific, social and professional worlds, this project has implied to take part to various manifestations which were national or regional, in the computer science domain or in others, scientific or non-scientific, organised by various political or non-political organisations working in the field of disabilities and accessibility. This phase has resulted in many reports which have directly fed into the bibliography of this thesis.

Orientador: Jose Roberto Sanches Mantovani / Resumo: Neste trabalho, propõe-se uma nova metodologia para abordar de forma integrada os problemas de restauração automática e sequenciamento de operação de abertura e fechamento de chaves em redes de distribuição de grande porte. Na literatura os problemas de restauração e sequência de chaveamentos são normalmente considerados de forma separada e sequencial, em que o resultado do algoritmo de restauração é o dado de entrada para o algoritmo que gera a sequência de chaveamento. A inconsistência com esta abordagem é que não necessariamente o resultado convencional do algoritmo de restauração (conjunto de chaves que devem ser manobradas), é o melhor dado de entrada para o algoritmo que elabora o sequenciamento ótimo de abertura/fechamento das chaves. Isso porque quando ambos os problemas são resolvidos separadamente, eles possuem funções objetivos diferentes. O problema de restauração tem por objetivo minimizar a quantidade de carga desconectada com o menor número de chaveamentos possíveis, enquanto que o problema de sequenciamento de chaves tem o objetivo de reduzir a energia não suprida no sistema durante um evento de falta permanente. Uma nova abordagem baseada na meta-heurística de Busca Tabu com Vizinhança Variável Reativa é proposta para explorar o espaço de busca do problema em análise, simultaneamente com uma nova heurística para gerar a sequência de chaveamento em sistemas de grande porte com milhares de nós de carga. A existência em operação na rede de controle de equipament... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: In this work, a new methodology is proposed to address, in an integrated approach, the automatic restoration problem and the switching sequence problem for large scale distribution networks. In the literature, the restoration and switching sequence problems are usually addressed separately and sequentially. Thus, the result of the distribution restoration algorithm is the initial data for the switching sequence algorithm. The inconsistency with this approach is that, not necessarily the conventional result of the restoration algorithm (a set of switches to be maneuvered) is the best initial data for the switching sequence algorithm. It is explained by the fact that both problems have different objective functions. The distribution restoration problem aims to minimize the amount of disconnected load with the fewest number of possible switching, whereas the switching sequence problem aims to minimize the energy not supplied in the network after a permanent fault. A new approach based on the Tabu Search with Reactive Variable Neighborhood meta-heuristic is proposed to explore the search space of the problem, along with a new heuristic to generate the switching sequence in large size distribution systems with thousands of load buses. The presence of voltage control devices, as switched capacitors and voltage regulators, are considered to improve the quality of solutions. The presence of distributed generation with black start capability is also considered. The cold load pick up c... (Complete abstract click electronic access below) / Doutor

O problema de dimensionamento e sequenciamento de lotes vem sendo extensivamente estudado por pesquisadores da área de Pesquisa Operacional e há uma tendência de que tais trabalhos passem a cada vez mais integrar aspectos reais dos processos produtivos. Entretanto, percebe-se que os estudos conduzidos em alguns setores industriais negligenciam importantes restrições tecnológicos dos processos de produção e isso afasta esses trabalhos de Pesquisa Operacional de uma aplicação efetiva, como é o caso da indústria de embalagens de vidro. Neste contexto, propõe-se um modelo de programação inteira mista e um método de solução para o problema de dimensionamento e sequenciamentos de lotes na indústria de embalagens de vidro, sendo que este trabalho diferencia-se dos demais existentes na literatura por agregar restrições tecnológicas específicas desse processo produtivo. O modelo proposto, denominado CLSD-GCST, foi amplamente validado com base em um conjunto de testes com 40 instâncias de um problema real de uma grande empresa do setor no pacote comercial IBM ILOG CPLEX Optimization Studio Versão 12.5. A validação do modelo incluiu ainda uma análise de ganhos potenciais para o negócio de baseada no modelo SCOR. Já o método de solução proposto consiste em uma metaheurística de Busca em Vizinhança Variável (VNS) e se mostrou promissor para a solução do problema estudado, proporcionando resultados de qualidade em um baixo tempo computacional. Além disso, o VNS superou o Branch-and-Cut do CPLEX para grandes instâncias, nas quais o pacote comercial encontrou dificuldades. Por fim, o VNS proposto também foi validado por meio da análise de testes computacionais e suas principais características foram avaliadas sistematicamente, gerando um conjunto de informações que pode direcionar a utilização e, até mesmo, a evolução desse método em pesquisas futuras. / Lot sizing and scheduling problem has been extensively studied by Operations Research scientists and there is a tendency of incorporating more production processes real aspects in these researches. However, it can be noticed that studies conducted in some industrial sectors neglect important production process technological constraints and it keeps the Operations Research works away from an effective application, as happens with the glass containers industry. In this context, a mixed integer programming model and a solution method were proposed for glass containers industry lot sizing and scheduling problem, the main difference between this work and the others in literature is the inclusion of process specific technological constraints. The proposed model, named CLSD-GCST, was widely validated by a set of tests performed with 40 instances from a large company real problem using the commercial package IBM ILOG CPLEX Optimization Studio Version 12.5. The model validation also included a potential business earnings analysis based on SCOR framework. About the proposed solution method, it consists of a Variable Neighborhood Search (VNS) metaheuristic and it proved to be promising for the studied problem solution, providing good quality results in low computational time. Moreover, VNS overcame the CPLEX Branch-and-Cut for large instances, in which the commercial package found difficulties. Lastly, the proposed VNS was validated by means of computational tests analysis and its main characteristics were systematically evaluated, generating an information set that may direct this method application and even its evolution in future researches.