Comparação de duas meta-heurísticas para a resolução do Problema de Roteamento de Veículos (PRV)

• Autor(a) principal: Morelli, Laura Caligaris Perim
• Data de Publicação: 2021
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da UFU
• Texto Completo: https://repositorio.ufu.br/handle/123456789/31794
• topic: Pesquisa operacional; Meta-heurística; PRV; VNS; ILS; Logística; CNPQ::ENGENHARIAS

The organizations search ways to improve their operations to maintain the high level of competitiveness, making possible the study and development of optimization models, where the vehicle routing fits on the classical problems from logistics management. The Vehicle Routing Problems (VRP) is an extension from the Travel Salesman Problem (TSP) when adding various restrictions. The VRP is classified as an NP-hard problem, in other words, there is no algorithm that finds an ideal solution in a polynomial time. In that context, the organizations with more optimized processes and low costs, survive the competition. This way, the VRP is one of the biggest problems to the routing for organizations and it fits on the classical problems from the logistics management. The present work is about a study where it is necessary to develop and implement a mathematical model on the optimization software Gurobi (2021) capable to determine routes to different vehicles with distinct capabilities, which will return an optimum solution indicating the best vehicle routing to be implemented, but much time is needed to return the solution. In order to speed the process, it was developed a constructive heuristic and, later, two metaheuristics: Variable Neighborhood Search (VNS) and Iterated Local Search (ILS) to compare the results between itself and seek the minimum cost from logistics operation by a distributor located in Guarulhos. The result was possible to develop a methodology to optimize the Vehicle Routing Problems (VRP) resolution, but it is possible to make improvements with a new local search.