Caminho mínimo com restrição probabilística de atraso máximo

In the Probabilistic Delay Constrained Shortest Path problem we aim to consider the time factor in the design of cargo routing paths in road networks at minimum cost, considering the increasing uncertainty in travel times of these routes in real networks, and keeping in mind strategies of quality of...

ver descrição completa

Detalhes bibliográficos
Autor: Araruna, Arthur Rodrigues
Formato: tesis de maestría
Estado:Versión publicada
Fecha de publicación:2013
País:Brasil
Recursos:Universidade Federal do Ceará (UFC)
Repositorio:Repositório Institucional da Universidade Federal do Ceará (UFC)
Idioma:portugués
OAI Identifier:oai:repositorio.ufc.br:riufc/17331
Acesso em linha:http://www.repositorio.ufc.br/handle/riufc/17331
Access Level:acceso abierto
Palavra-chave:Matematica discreta e combinatória
Caminho mínimo
Restrição probabilística
Otimização inteira
Shortest path
Probabilistic constraint
Integer optimization
Descrição
Resumo:In the Probabilistic Delay Constrained Shortest Path problem we aim to consider the time factor in the design of cargo routing paths in road networks at minimum cost, considering the increasing uncertainty in travel times of these routes in real networks, and keeping in mind strategies of quality of service, in order to obtain a compromise between the travel costs and the compliance of the arrival time at the destination. We conducted a study of related problems in the literature of transport networks optimization, in order to better understand the problem to be addressed, about which we are not aware of existing works. We developed a scheme for enumerating partitions of the solution space of this problem, which uses an L decomposition to select these partitions wisely, and is aided by solutions to relaxations of the problem to obtain bounds for the optimal cost. In addition, we developed some branching and pruning strategies for a Branch-and-Bound scheme, with a pre-processing phase, in order to try and solve the problem directly. The computational results show that we are competitive with the commercial tool used for comparison in the smaller instances. For the remaining instances, this tool is more efficient in the time required for solving the problem.