Metaheurística variable neighborhood search (VNS) e variable neighborhood descent (VND) aplicada na distribuição de combustíveis em rede multimodal
This work presents a mathematical model for the problem of petroleum products distribution in multimodal network as a particular case of the transportation problem known in the literature as Two-echelon Capacitated Vehicle Routing Problem (2E-CVRP), which performs vehicle routing in two levels, wher...
| Autor principal: | Oliveira, João Wagner de |
|---|---|
| Formato: | Dissertação |
| Idioma: | Português |
| Publicado em: |
Universidade Tecnológica Federal do Paraná
2021
|
| Assuntos: | |
| Acesso em linha: |
http://repositorio.utfpr.edu.br/jspui/handle/1/24729 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
| Resumo: |
This work presents a mathematical model for the problem of petroleum products distribution in multimodal network as a particular case of the transportation problem known in the literature as Two-echelon Capacitated Vehicle Routing Problem (2E-CVRP), which performs vehicle routing in two levels, where vehicles leave the warehouses with a limited amount of cargo and transport them to an intermediate warehouse, where the product is divided and redistributed in new vehicles that leave for the customer to deliver. To solve this problem, a mathematical model was generated and solved using the Variable Neighborhood Search (VNS) metaheuristic, which uses concepts of mathematical optimization to perform searches from an initial solution in search of better solutions. To aid in the search, the Variable Neighborhood Descent (VND) heuristic was implemented. The method consists of defining a limited number of different neighborhoods to start the search for a better solution until all the neighborhoods are evaluated. The goal is to find the connection arcs between the refineries and the distribution centers and the distribution centers to the customers that minimize the objective function. The metaheuristic was able to find the optimal solution in a satisfactory computational time, proving to be effective in assisting decision making. The metaheuristic reached the result in 0.7 seconds, while the mathematical model, which was implemented in Excel and solved by Gurobi, reached the result in 0.2 seconds. |
|---|