Cette thèse est une contribution aux travaux de recherche sur l’optimisation des problèmes du transport multimodal. Les principaux concepts clé de la multimodalité dans les réseaux du transport intermodal et l’état de l’art des travaux scientifique du domaine y sont présentés. Le problème de la localisation des terminaux du transport combiné est ensuite étudié. Nous proposons un algorithme génétique à codage mixte pour la résolution de ce problème et nous comparons nos résultats avec ceux de la littérature. Un ensemble de problèmes posés dans le cadre de notre travail sur le projet DCAS (Direct Cargo Axe Seine), porté par le Grand Port Maritime du Havre, y est décrit et modélisé par des outils de programmation mathématique. Ainsi, nous avons étudié le problème du transfert de navettes ferroviaires qui consiste à optimiser le transfert d’un ensemble de conteneurs entre des terminaux maritimes et un terminal multimodal. Ensuite, nous avons modélisé le problème d’ordonnancement des trains de grandes de lignes pour le placement sur les voies de la cour ferroviaire du terminal multimodal du Havre. Ces problèmes sont résolus en utilisant une approche combinée optimisation-simulation. Une première application est basée sur un algorithme génétique couplé avec la simulation multi agents pour l’affectation des voies aux trains. Une deuxième, consiste à optimiser la manutention des conteneurs lors d’un transbordement rail-rail en utilisant un algorithme de colonie de fourmis intégré dans le modèle de simulation et une stratégie de collaboration agents pour minimiser les temps d’attente des portiques et ainsi augmenter leurs productivités. / This thesis is a contribution to research on the optimization of multimodal transport problems. The main key concepts of multimodality in the intermodal transportation networks and the state of the art of scientific works in the field are presented. The intermodal terminal location problem is then studied. We propose a genetic algorithm with mixed encoding for solving this problem and we compare our results with those of literature. A set of problems in the framework of our work on the project DCAS (Direct Cargo Axe Seine), carried by the Grand Port Maritime du Havre, are described and modeled by mathematical programming tools. Thus, we studied the problem of the transfer of rail shuttles which is to optimize the transfer of a set of containers between maritime terminals and a multimodal terminal. We then modeled the scheduling problem of freight trains for placement on rail tracks. These problems are solved by using combined optimization simulation approaches. A first application is based on a genetic algorithm coupled with the multi agent’s simulation. A second is to optimize a rail-rail transshipment of containers using an ant colony algorithm embedded in the simulation model and an agent’s collaboration strategy to minimize waiting times and increase cranes productivity.541 ##‎$a@Optimization of multimodal transportation problems

Les métaheuristiques sont des techniques d’optimisation indépendantes des problèmes traités. Elles ne profitent pas d’une spécificité du problème et, par conséquent, peuvent fournir des cadres généraux qui peuvent être appliqués a de nombreuses classes de problèmes. Les métaheuristiques peuvent fournir une stratégie de guidage dans la conception des heuristiques pour résoudre des problèmes d’optimisation spécifiques. Leur utilisation dans de nombreuses applications montre leur efficacité pour résoudre des problèmes importants et complexes. De nos jours, les métaheuristiques appliquées `a la solution des problèmes d’optimisation ont évolué vers l’intégration d’autres techniques d’optimisation, de sorte que les méthodes de résolution peuvent bénéficier des avantages de chacune des composantes. Le travail dans cette thèse vise à contribuer à l’étude des problèmes de tournées de véhicules avec des visites optionnelles en fournissant un opérateur à base de programmation dynamique intégré dans un processus métaheuristique générique. L’opérateur récupère le tour optimal de clients à visiter, répondant aux contraintes du problème, tout en optimisant l’objectif défini. L’opérateur pose le problème de la sélection des meilleurs clients `a visiter comme un problème de plus court chemin avec contraintes de ressources sur un graphe auxiliaire dirigé acyclique représentant les choix de visite possibles. Dans les problèmes de tournées de véhicules avec des visites optionnelles, les clients à servir ne sont pas connus a priori et cela rend plus difficile à résoudre le problème qu’un problème de routage classique qui est lui-même déjà NP-difficile. Les problèmes de tournées avec des visites optionnelles trouvent des applications dans des domaines multiples et variés tels que la conception de la distribution, la logistique humanitaire, la prestation des soins de santé, le tourisme, le recrutement, la collection ou la livraison de marchandises et patrouille en milieu urbain / Metaheuristics are problem independent optimisation techniques. As such, they do not take advantage of any specificity of the problem and, therefore, can provide general frameworks that may be applied to many problem classes. These iterative upper level methodologies can furnish a guiding strategy in designing subordinate heuristics to solve specific optimisation problems. Their use in many applications shows their efficiency and effectiveness to solve large and complex problems. Nowadays, metaheuristics applied to the solution of optimisation problems have shifted towards integrating other optimisation techniques, so that solution methods benefit from the advantages each offers. This thesis seeks to contribute to the study of vehicle routing problems with optional visits by providing a dynamic programming-based operator that works embedded into a generic metaheuristic. The operator retrieves the optimal tour of customers to visit, satisfying the side constraints of the problem, while optimising the defined objective. The operator formulates the problem of selecting the best customers to visit as a Resource Constrained Elementary Shortest Path Problem on an auxiliary directed acyclic graph where the side restrictions of the problem considered act as the constraining resource. In vehicle routing problems with optional visits, the customers to serve are not known a priori and this fact leaves a more difficult to solve problem than a classic routing problem, which per se is already NP-hard. Routing problems with optional visits find application in multiple and diverse areas such as bimodal distribution design, humanitarian logistics, health care delivery, tourism, recruitment, hot rolling production, selected collection or delivery, and urban patrolling among others

La segmentation d'image est le processus de partitionnement d'une image numérique en régions, non chevauchées, homogènes vis-à-vis de certaines caractéristiques, telles que le niveau de gris, la texture, le mouvement, etc. Elle a des applications dans plusieurs domaines comme l'imagerie médicale, la détection d'objets, la biométrie, l'imagerie par satellite, la navigation de robot, la vidéosurveillance, etc. Le processus de segmentation représente une étape cruciale dans les systèmes de vision par ordinateur, car les caractéristiques et décisions sont extraites et prises à partir de son résultat. Les premiers algorithmes de segmentation d'image ont vu le jour dans les années 1970. Depuis, de nombreuses techniques et méthodes de segmentation ont été expérimentées pour essayer d'améliorer les résultats. Néanmoins, jusqu'à nos jours, aucun algorithme de segmentation d'image n'arrive à fournir des résultats parfaits sur une large variété d'images. Les "métaheuristiques" sont des procédures conçues pour résoudre des problèmes d'optimisation dits difficiles. Ce sont en général des problèmes aux données incomplètes, incertaines, bruitées ou confrontés à une capacité de calcul limitée. Les métaheuristiques ont connu un succès dans une large variété de domaines. Cela découle du fait qu'elles peuvent être appliquées à tout problème pouvant être exprimé sous la forme d'un problème d'optimisation de critère(s). Ces méthodes sont, pour la plupart, inspirées de la physique (recuit simulé), de la biologie (algorithmes évolutionnaires) ou de l'éthologie (essaims particulaires, colonies de fourmis).Ces dernières années, l'introduction des métaheuristiques dans le domaine du traitement d'images a permis d'étudier la segmentation sous un angle différent, avec des résultats plus ou moins réussis. Dans le but d'apporter notre contribution et d'améliorer davantage les performances des méthodes de segmentation, nous avons proposé des algorithmes basés régions, contours et hybrides, mettant en œuvre des métaheuristiques d'optimisation dans des approches mono et multiobjectif. Les méthodes proposées ont été évaluées sur des bases de données expérimentales composées d'images synthétiques, d'images IRM simulées et d'images IRM réelles ainsi que des images de tomographie par émission de positons (TEP). Les résultats obtenus sont significatifs et prouvent l'efficacité des idées proposées / Image segmentation is the process of partitioning a digital image into homogeneous non-overlapped regions with respect to some characteristics, such as gray value, motion, texture, etc. It is used in various applications like medical imaging, objects detection, biometric system, remote sensing, robot navigation, video surveillance, etc. The success of the machine vision system depends heavily on its performance, because characteristics and decisions are extracted and taken from its result. The first image segmentation algorithms were introduced in the 70's. Since then, various techniques and methods were experimented to improve the results. Nevertheless, up till now, no method produces a perfect result for a wide variety of images. Metaheuristics are a high level procedure designed to solve hard optimization problems. These problems are in general characterized by their incomplete, uncertain or noised data, or faced to low computing capacity. Metaheuristics have been extremely successful in a wide variety of fields and demonstrate significant results. This is due to the fact that they can applied to solve any problem which can be formulated as an optimization problem. These methods are, mainly, inspired from physics (simulated annealing), biology (evolutionary algorithms), or ethology (particle swarm optimization, ant colony optimization).In recent years, metaheuristics are starting to be exploited to solve segmentation problems with varying degrees of success and allow to consider the problem with different perspectives. Bearing this in mind, we propose in this work three segmentation and post-segmentation approaches based on mono or multiobjective optimization metaheuristics. The proposed methods were evaluated on databases containing synthetic images, simulated MRI images, real MRI images and PET images. The obtained results show the efficiency of the proposed ideas

Cette thèse porte sur la résolution de problèmes de transport qui intègrent des contraintes temporelles considérant les fenêtres de temps, la synchronisation des visites et l’équilibrage des services. Ces problèmes trouvent plusieurs applications dans le monde réel.L’objectif de nos recherches est l’élaboration de nouvelles méthodes de résolution pour les problèmes considérés en examinant leur performance avec une étude comparative par rapport aux différentes approches de la littérature. Deux variantes sont traitées. Le premier cas étudie le Problème de Tournées de Véhicules avec Fenêtres de Temps (VRPTW). Nous proposons de nouveaux prétraitements et bornes inférieures pour déterminer le nombre de véhicules nécessaires en s’inspirant de travaux menés en ordonnancement (raisonnement énergétique) et d’autres problèmes combinatoires comme la clique maximum et les problèmes de bin-packing. Nous présentons également un algorithme d’optimisation par essaim particulaire qui traite de la minimisation du nombre de véhicules puis de celle du temps de trajet total. Le deuxième cas étudie le Problème de Tournées de Véhicules avec des Fenêtres de Temps et des Visites Synchronisées (VRPTWSyn). Nous proposons plusieurs méthodes basées sur des approches heuristiques et des formulations linéaires avec l’incorporation d’inégalités valides pour tenir compte de la contrainte de synchronisation. / This dissertation focuses on vehicle routing problems, one of the major academic problems in logistics. We address NP-Hard problems that model some realworld situations particularly those with different temporal constraints including time windows, visit synchronization and service balance.The aim of this research is to develop new algorithms for the considered problems,investigate their performance and compare them with the literature approaches.Two cases are carried out. The first case studies the Vehicle Routing Problem with Time Windows (VRPTW). We propose new lower bound methods for the number of vehicles. Then we present a Particle Swarm Optimization algorithm dealing with the Solomon objective. The second case studies the VehicleRouting Problem with Time Windows and Synchronized Visits (VRPTWsyn).Both exact methods and heuristics are proposed and compared to the literature approaches.

O presente documento tem por objetivo apresentar o desenvolvimento de uma ferramenta computacional para auxiliar profissionais da área de otimização na implementação de métodos e resolução de problemas. O projeto foi desenvolvido como tema de dissertação no Programa de Mestrado em Ciência da Computação e Matemática Computacional do ICMC/USP. A ferramenta pode ser enquadrada como um ambiente de desenvolvimento (framework) e será chamada de ProOF - Professional Optimization Framework. O ProOF tem como foco principal nortear a implementação computacional de métodos variados para problemas de otimização, utilizando como paradigma a programação orientada a objetos. Esse framework incorpora as principais características encontradas por outras ferramentas propostas na literatura. Além disso, procura facilitar a implementação de métodos e resolução de problemas ao permitir alto reuso de códigos, dar suporte a geração de códigos em diferentes linguagens de programação e gerar uma Graphical User Interface (GUI) automática para parametrização dos métodos inseridos pelo usuário. Alguns trabalhos publicados recentemente utilizaram versões em desenvolvimento do ProOF e serão citados como estudo de caso para atestar a robustez do framework proposto. Por fim, uma comparação será realizada entre o ProOF e outros frameworks existentes na literatura / This paper aims to present the development of a computational tool to assist professionals in the optimization field in implementation of methods and problem solving. The project was developed as dissertation topic in the Masters Program in Computer Science and Computational Mathematics at ICMC/USP. The tool can be considered as a development environment (framework) and will be called ProOF - Professional Optimization Framework. The ProOF is mainly focused on guiding the implementation of various computational methods for optimization problems using as a paradigm the object-oriented programming. This framework incorporating the principal features found in other tools proposed in the literature. Moreover, seeks to facilitate the implementation of methods and problem resolution by allowing high code reuse, give support to code generation in different programming languages and generate a Graphical User Interface (GUI) automatic for parameter setting of methods implemented by the user. Some recently published studies have used previous versions of the ProOF and they will be cited as a case study to attest the robustness of the proposed framework. Finally, a comparison will be made between the ProOF and other existing frameworks in the literature

Deux problèmes robustes d'optimisation NP-difficiles sont étudiés dans cette thèse: le problème min-max regret de couverture pondérée (min-max regret WSCP) et le problème min-max regret de couverture et localisation maximale (min-max regret MCLP). Les données incertaines dans ces problèmes sont modélisées par des intervalles et seules les valeurs minimales et maximales pour chaque intervalle sont connues. Le min-max regret WSCP a été investigué notamment dans le cadre théorique, alors que le min-max regret MCLP a des applications en logistique des catastrophes étudiées dans cette thèse. Deux autres critères d'optimisation robuste ont été dérivés pour le MCLP: le max-max MCLP et le min-max MCLP. En matière de méthodes, formulations mathématiques, algorithmes exacts et heuristiques ont été développés et appliqués aux deux problèmes. Des expérimentations computationnelles ont montré que les algorithmes exacts ont permis de résoudre efficacement 14 des 75 instances générées par le min-max regret WSCP et toutes les instances réalistes pour le min-max regret MCLP. Pour les cas simulés qui n'ont pas été résolus de manière optimale dans les deux problèmes, les heuristiques développées dans cette thèse ont trouvé des solutions aussi bien ou mieux que le meilleur algorithme exact dans presque tous les cas. En ce qui concerne l'application en logistique des catastrophes, les modèles robustes ont trouvé des solutions similaires pour les scénarios réalistes des tremblements de terre qui a eu lieu à Katmandu au Népal en 2015. Cela indique que nous avons une solution robuste / Two robust optimization NP-Hard problems are studied in this thesis: the min-max regret Weighted Set Covering Problem (min-max regret WSCP) and the min-max regret Maximal Coverage Location Problem (min-max regret MCLP). The min-max regret WSCP and min-max regret MCLP are, respectively, the robust optimization counterparts of the Set Covering Problem and of the Maximal Coverage Location Problem. The uncertain data in these problems is modeled by intervals and only the minimum and maximum values for each interval are known. However, while the min-max regret WSCP is mainly studied theoretically, the min-max regret MCLP has an application in disaster logistics which is also investigated in this thesis. Two other robust optimization criteria were derived for the MCLP: the max-max MCLP and the min-max MCLP. In terms of methods, mathematical formulations, exact algorithms and heuristics were developed and applied to both problems. Computational experiments showed that the exact algorithms efficiently solved 14 out of 75 instances generated to the min-max regret WSCP and all realistic instances created to the min-max regret MCLP. For the simulated instances that was not solved to optimally in both problems, the heuristics developed in this thesis found solutions, as good as, or better than the exact algorithms in almost all instances. Concerning the application in disaster logistics, the robust models found similar solutions for realistic scenarios of the earthquakes that hit Kathmandu, Nepal in 2015. This indicates that we have got a robust solution, according to all optimization models

Metaheuristics is a term for optimization procedures/algorithms that can be applied to a wide range of problems. These problems for which metaheuristics are used usually fall in the NP-hard category, meaning that they cannot be solved in polynomial time. This means that as the input dataset gets larger the time to solve increases exponentially. One such problem is the traveling salesman problem (TSP) which is and has been widely used as a benchmark problem to test optimization algorithms. This study focused on two such algorithms called ant colony optimization (ACO) and genetic algorithm (GA) respectively. Development of such optimization algorithms can have huge implications in several areas of business and industry. They can for example be used by delivery companies to optimize routing of delivery vehicles as well as in material science/industry where they can be used to calculate the most optimal mix of ingredients to produce materials with the desired characteristics. The approach taken in this study was to compare the performance of the two algorithms in three different programming languages (python, javascript and C#).  Previous studies comparing the two algorithms have reported conflicting results where some studies found that ACO yielded better results but was slower than GA, while others found that GA yielded better results than ACO. Results of this study suggested that both ACO and GA could find the benchmark solution, but  ACO did so much more consistently. Furthermore javascript was found to be the most efficient language with which to run the algorithms in the setup used in this study.

In countries like the US, public school systems function through school districts, which are geographical areas where schools share the same administrative structure and are often coterminous with the boundary of a city or a county. School districts play an important role in the functioning of society. In a well-run school district with safe and well-functioning schools, graduating enough students can enhance the quality of life in its area. Conversely, a poorly run district may cause growth in the area to be far less than surrounding areas, or even a decline in population over time. To promote the efficient functioning of the school district, the boundaries of public schools are redrawn from time to time by the school board/planning officials. In the majority of the cases, this process of redrawing the school boundaries, also called school redistricting or school boundary formation, is done manually by the planners and involves hand-drawn maps. Given the rapid advancements in GIS made in the last decade and the availability of high-quality geospatial data, we opine that an objective treatment of the school redistricting problem by a data-driven model can assist the school board/ decision-makers by providing them with automated plans. These automated plans may serve as possible suggestions to the planners, who can adapt them to prepare their own plans in the way they see fit based on their subjective knowledge and expertise. In this dissertation, we propose algorithmic techniques for solving the problem of (school) redistricting, which is an NP-hard problem. We primarily investigate optimization-based algorithms for solving the problem. Our approaches include (i) clustering, (ii) local search, and (iii) memetic algorithms. We also propose ways of solving the problem using exact methods and fair redistricting techniques based on ethical considerations. The techniques developed here are generic enough to be applied to other redistricting problems with some degree of modification in the objective function and constraint-handling techniques. The source code and corresponding datasets are available at https://github.com/subhodipbiswas/schoolredistricting. / Doctor of Philosophy / In many countries, public school systems function through school districts, which are geographical areas where schools share the same administrative structure and are often coterminous with the boundary of a city or a county. To promote efficient functioning of the school district, the boundaries of public schools are redrawn from time to time by the school board/planning officials. In the majority of the cases, this process of redrawing the school boundaries, also called school redistricting, is done manually by the planners and involves hand-drawn maps. Given the rapid advancements in GIS made in the last decade and the availability of high-quality geospatial data, we opine that an objective treatment of the school redistricting problem by a data-driven model can assist the school board/ decision-makers by providing them with automated plans. In this presentation, we propose algorithmic techniques for solving the school redistricting problem. Our approaches include (i) clustering, (ii) local search, and (iii) memetic algorithms. We also show that MCMC-based techniques can aid in enabling exact methods to operate on this problem. Lastly, we briefly highlight ethical considerations involved in the process of school redistricting and throw light on some ways to devise more ethically-aware strategies for doing school redistricting. The results indicate that the proposed methods could be a valuable decision-making tool for school officials.

<p> In this thesis, we have considered the vehicle routing problem with pickup and delivery which is a generalization of the capacitated vehicle routing problem (CVRP). The vehicle routing problem with pickup and delivery (VRPPD) arises whenever pickup demand and delivery demand is to be satisfied by the same vehicle. The problem is encountered in many real life situations including reverse logistics. We consider three variants of VRPPD, namely, the vehicle routing problem with back-hauls (VRPB), the vehicle routing problem with back-hauls and mixed-load (VRPBM) and the vehicle routing problem with simultaneous pickup and delivery (VRPSPD). </p> <p> The inherent complexity of VRPPD makes it an NP -hard problem. It is not possible to solve an NP-hard problem in polynomial time unless P = NP. Therefore, heuristics and metaheuristics are used to produce a good quality solution within reasonable CPU time. We develop ant colony algorithms for VRPB, VRPBM and VRPSPD. We have improved the existing ant-colony algorithms by applying better local search schemes and by adding new features such as construction rule and the trail updating criteria. We also develop saving based heuristics for single and multi-depot versions of VRPSPD. Checking feasibility of a given route is an important issue in VRPSPD because of the fluctuating load on the vehicle. We have proposed the cumulative net-pickup approach for this purpose. One important feature of this approach is that it checks the feasibility of an altered route in constant time. </p> <p> The proposed heuristics and metaheuristics are evaluated by solving benchmark problem instances available in literature and then comparing the solutions with the solutions produced by the existing algorithms. Our computational experiment has shown that the proposed heuristics and metaheuristics give better or equally good results in comparison to the existing solution procedures. </p> / Thesis / Doctor of Philosophy (PhD)

Hundreds of steel products need to be scheduled on a single or parallel machine in a steel plant every week. A good feasible schedule may save the company millions of dollars compared to a bad one. Single and parallel machine scheduling are also encountered often in many other industries, making it a crucial research topic for both the process system engineering and operations research communities. Single or parallel machine scheduling can be a challenging combinatorial optimization problem when a large number of jobs are to be scheduled. Each job has unique job characteristics, resulting in different setup times/costs depending on the processing sequence. They also have specific release dates to follow and due dates to meet. This work presents both an exact method using mixed-integer quadratic programming, and an approximate method with metaheuristics to solve real-world large-scale single/parallel machine scheduling problems faced in a steel plant. More than 1000 or 350 jobs are to be scheduled within a one-hour time limit in the single or parallel machine problem, respectively. The objective of the single machine scheduling is to minimize a combined total changeover, total earliness, and total tardiness cost, whereas the objective of the parallel machine scheduling is to minimize an objective function comprising the gaps between jobs before a critical time in a schedule, the total changeover cost, and the total tardiness cost. The exact method is developed to benchmark computation time for a small-scale single machine problem, but is not practical for solving the actual large-scale problem. A metaheuristic algorithm centered on variable neighborhood descent is developed to address the large-scale single machine scheduling with a sliding-window decomposition strategy. The algorithm is extended and modified to solve the large-scale parallel machine problem. Statistical tests, including Student's t-test and ANOVA, are conducted to determine efficient solution strategies and good parameters to be used in the metaheuristics. / Thesis / Master of Applied Science (MASc)