The Vehicle Routing Problem with Floating Targets: Formulation and Solution Approaches
This paper addresses a generalization of the vehicle routing problem in which the pick-up locations of the targets are nonstationary. We refer to this problem as the vehicle routing problem with floating targets and the main characteristic is that targets are allowed to move from their initial home...
| Autores: | , , |
|---|---|
| Formato: | artículo |
| Fecha de publicación: | 2018 |
| País: | España |
| Recursos: | IE |
| Repositorio: | Repositorio IE |
| OAI Identifier: | oai:repositorio.ie.edu:20.500.14417/4110 |
| Acesso em linha: | https://doi.org/10.1287/ijoc.2017.0800 https://hdl.handle.net/20.500.14417/4110 https://pubsonline.informs.org/doi/10.1287/ijoc.2017.0800 |
| Access Level: | acceso abierto |
| Palavra-chave: | 33 Ciencias Tecnológicas::3307 Tecnología electrónica ODS 7 - Energía asequible y no contaminante |
| Resumo: | This paper addresses a generalization of the vehicle routing problem in which the pick-up locations of the targets are nonstationary. We refer to this problem as the vehicle routing problem with floating targets and the main characteristic is that targets are allowed to move from their initial home locations while waiting for a vehicle. This problem models new applications in drone routing, ridesharing, and logistics where a vehicle agrees to meet another vehicle or a customer at a location that is away from the designated home location. We propose a Mixed Integer Second Order Cone Program (MISOCP) formulation for the problem, along with valid inequalities for strengthening the continuous relaxation. We further exploit the problem structure using a Lagrangian decomposition and propose an exact branch-and-price algorithm. Computational results on instances with varying characteristics are presented and the results are compared to the solution of the full problem using CPLEX. The proposed valid inequalities reduce the computational time of CPLEX by up to 30% on average while the proposed branch and price is capable of solving instances where CPLEX fails in finding the optimal solution within the imposed time limit. |
|---|