A carregar...
Publicação
Comparação de modelos para o Fleet Assignment Problem (Problema de Afetação de Frotas)
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.44 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Na indústria da aviação é crucial a gestão minuciosa em todos os campos. Dada a complexidade de otimização das operações ligadas ao voo, o planeamento de uma companhia aérea é usualmente dividido em diversas fases sequenciais: planeamento da frota, planeamento das rotas a operar, afetação da frota, gestão e manutenção da mesma e afetação da tripulação. Nesta dissertação aborda-se o problema de afetação de frota de forma isolada, embora alguns autores optem por integrar este sub-problema com considerações mais amplas relativas à manutenção dos aparelhos e/ou afetação da tripulação. Considerando um conjunto de voos que uma dada companhia aérea (com um hub fixo) tem agendados, e conhecendo as estimativas de procura associada a esses mesmos voos, torna-se perentório utilizar a frota disponível da melhor forma, de modo a que a cada voo seja afeto o avião mais adequado em termos de capacidade e minimizando os custos associados. Consideram-se dois modelos com estrutura de redes distintas, que serão comparados relativamente à dimensão, tempo de execução e qualidade dos limites inferiores obtidos através da relaxação linear, face a um conjunto de instâncias. A aplicação dos modelos será ainda estendida a um horizonte temporal diferente do apresentado na literatura. Para cada instância foram testados cenários com diferentes níveis de procura e com diferentes composições de frota. As soluções obtidas com estes cenários foram comparadas, indicando as melhores opções para cada caso.
In the aviation industry, the thorough management in all fields is crucial. Given the complexity of flight operations optimization, the planning of an aviation company is usually divided in several sequential stages: the fleet planning, the routes planning, the fleet assignment, the aircraft routing and management and the crew scheduling. In this thesis, the fleet assignment problem is approached in an isolated way, although some authors choose to integrate this subproblem in a framework considering factors like aircraft maintenance and crew scheduling. Given a set of scheduled flights of some aviation company with a fixed hub, and knowing the associated flights demand, it becomes peremptory the good usage of the available fleet in order to assign to each flight the most suitable aircraft in terms of capacity, minimizing at the same time the associated costs. Two models are considered with different network structures that will be compared against a set of instances in terms of the size, execution time and the quality of the lower bounds obtained by the linear relaxation. The models application will be extended to a time horizon different from that presented in the literature. For each instance different demand level scenarios and different fleet composition scenarios were tested. The solutions obtained with these scenarios were compared, indicating the best options for each case.
In the aviation industry, the thorough management in all fields is crucial. Given the complexity of flight operations optimization, the planning of an aviation company is usually divided in several sequential stages: the fleet planning, the routes planning, the fleet assignment, the aircraft routing and management and the crew scheduling. In this thesis, the fleet assignment problem is approached in an isolated way, although some authors choose to integrate this subproblem in a framework considering factors like aircraft maintenance and crew scheduling. Given a set of scheduled flights of some aviation company with a fixed hub, and knowing the associated flights demand, it becomes peremptory the good usage of the available fleet in order to assign to each flight the most suitable aircraft in terms of capacity, minimizing at the same time the associated costs. Two models are considered with different network structures that will be compared against a set of instances in terms of the size, execution time and the quality of the lower bounds obtained by the linear relaxation. The models application will be extended to a time horizon different from that presented in the literature. For each instance different demand level scenarios and different fleet composition scenarios were tested. The solutions obtained with these scenarios were compared, indicating the best options for each case.
Descrição
Tese de mestrado, Estatística e Investigação Operacional (Investigação Operacional) Universidade de Lisboa, Faculdade de Ciências, 2019
Palavras-chave
Afetação de Frota Redes Aéreas Otimização Problema Inteiro Relaxação Linear Teses de mestrado - 2019