Descripción
|
|
---|---|
The design of tourist itineraries is one of the many practical applications of the NP-hard problem known as the Generalized Orienteering Problem (GOP). The GOP extends the well-known Orienteering Problem (OP) by dealing with multiple-scored attractions and a nonlinear objective function. As in the original OP, a set of nodes that could potentially be visited is given and the travel time between any pair of nodes is known, together with the time budget. However, the difference with the OP is that in the GOP, each node is associated with several scores, and the objective consists of finding a closed tour maximizing a weighted sum of different score types. Due to its non-linear objective function, the GOP has been approached using different metaheuristics, including Neural Networks, Genetic Algorithms, and others. In this work, we propose a Variable Neighborhood Search (VNS) to solve it. Our VNS uses a reduced number of local search operators and performs the calculation of the scores in an e?fficient way. In the literature, a case study of 27 Chinese cities was used as a benchmark by most of the authors approaching the GOP, so we also use it to evaluate the performance of our algorithm. Furthermore, we have also created some more data sets to test the performance of the VNS. In the experiments, the VNS has been able to find better local optima in a shorter computational time in most cases. | |
Internacional
|
Si |
Nombre congreso
|
29th European Conference on Operational Research |
Tipo de participación
|
960 |
Lugar del congreso
|
Valencia, España |
Revisores
|
No |
ISBN o ISSN
|
978-84-09-02938-9 |
DOI
|
|
Fecha inicio congreso
|
08/07/2018 |
Fecha fin congreso
|
11/07/2018 |
Desde la página
|
142 |
Hasta la página
|
142 |
Título de las actas
|
Proceedings of the 29th European Conference on Operational Research |