Made available in DSpace on 2014-12-17T15:26:37Z (GMT). No. of bitstreams: 1 JoaoSSN.pdf: 5224762 bytes, checksum: 4021177e0509af10223ad40751ece2f0 (MD5) Previous issue date: 2009-07-10 / The present essay shows strategies of improvement in a well succeded evolutionary metaheuristic to solve the Asymmetric Traveling Salesman Problem. Such steps consist in a Memetic Algorithm projected mainly to this problem. Basically this improvement applied optimizing techniques known as Path-Relinking and Vocabulary Building. Furthermore, this last one has being used in two different ways, in order to evaluate the effects of the improvement on the evolutionary metaheuristic. These methods were implemented in C++ code and the experiments were done under instances at TSPLIB library, being possible to observe that the procedures purposed reached success on the tests done / O presente trabalho prop?e estrat?gias de melhoria em uma bem sucedida metaheur ?stica evolucionaria para a resolu??o do Problema do Caixeiro Viajante Assim?trico. Tal procedimento consiste em um algoritmo mem?tico projetado especificamente para esse problema. Essas melhorias t?m por base a aplica??o de t?cnicas de otimiza??o conhecidas como Path-Relinking e Vocabulary Building, sendo essa ?ltima t?cnica utilizada de dois modos distintos, com o intuito de avaliar os efeitos de melhoria sobre a metaheur?stica evolucion?ria empregada. Os m?todos propostos foram implementados na linguagem de programa??o C++ e os experimentos computacionais foram realizados sobre inst?ncias disponibilizadas na biblioteca TSPLIB, tornando poss?vel observar que os procedimentos propostos alcan?aram ?xito nos testes realizados

Notre objectif dans cette thèse est de proposer des algorithmes efficaces pour résoudre des problèmes d’optimisation combinatoire difficiles. Dans un premier temps, nous établissons le principe de l’énumération ordonnée qui consiste à générer dans un ordre adéquat les solutions d’un problème relâché associé au problème principal jusqu’à l’obtention de la preuve d’optimalité d’une solution. Nous construisons une procédure générique dans le cadre général des problème d’optimisation combinatoire. Dans un second temps nous abordons les applications de notre algorithme sur des problèmes qui admettent le problème d’affectation comme relaxation. Le premier cas particulier que nous étudions est la recherche d’une solution de bon compromis pour le problème d’affectation multiobjectif. La seconde application se rapporte au problème du voyageur de commerce asymétrique qui présente la difficulté de comporter des contraintes qui interdisent les sous-tournées, en plus des contraintes du problème d’affectation. / Our aim in this thesis is to propose efficient algorithms for solving difficult combinatorial optimization problems. Our algorithms are based on a generic method of ordered enumeration. Initially, we describe the principle of ordered enumeration which consists in generating in a specific order solutions of a relaxed problem associated to the difficult main problem, until meeting a proof of the optimality of a feasible solution. We construct a generic procedure in the general context of combinatorial optimization problems. In a second step we discuss applications of our algorithm on some difficult problems which admit the assignment problem as relaxation. The first special case we study is the search for a compromise solution to the multiobjective assignment problem. The second application is the asymmetric travelling salesman problem, which contains sub-tour constraints in addition to the constraints of the assignment problem.