Uso del operador swap genera soluciones eficientes computacionales en un caso de enrutamiento de vehículos con enfoque de ventanas de tiempo
Introduction— Vehicle routing scheduling with service compliance is a necessity for logistics companies in search of their competitive advantage. Objective— The objective of the following work is to determine the routing of vehicles with time windows for a homogeneous fleet applied to the last, mile...
| Autor: | |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión aceptada para publicación |
| Fecha de publicación: | 2021 |
| País: | Colombia |
| Institución: | Corporación Universidad de la Costa |
| Repositorio: | Repositorio REDICUC |
| Idioma: | español |
| OAI Identifier: | oai:repositorio.cuc.edu.co:11323/8742 |
| Acceso en línea: | https://hdl.handle.net/11323/8742 https://doi.org/10.17981/cesta.02.01.2021.05 https://repositorio.cuc.edu.co/ |
| Access Level: | acceso abierto |
| Palabra clave: | VRP applications Heuristic Saving matrix Combinatorial optimization VRPTW Aplicaciones VRP Heurística Matriz de ahorro Optimización combinatoria |
| Sumario: | Introduction— Vehicle routing scheduling with service compliance is a necessity for logistics companies in search of their competitive advantage. Objective— The objective of the following work is to determine the routing of vehicles with time windows for a homogeneous fleet applied to the last, mile distribution case with 300 clients, considering the minimization of operating costs, distribution costs and, downtime costs. Methodology— The problem is approached through the approach of a mixed-integer linear programming mathematical model, and the development of an algorithm through the use of the savings method and the use of the swap operator. Results— In the construction phase, the savings algorithm achieves an initial cost focused on the minimum distance. In the upgrade phase, the swap operator improves the initially established solution, very quickly. For a case of 300 clients, 12 iterations were carried out, obtaining an improvement of 71.41% over the initial cost. Conclusions— For calculations of VRPTW cases with 300 nodes, the swap operator achieves computational times of less than 30 seconds. |
|---|