Long Term Interval Demand Forecasting for Air Transportation Networks
- Bachir Amadou
- Luis Gustavo Zelaya Cruz
- Mohamed Sbihi
- Daniel Delahaye
Universidade Federal Fluminense, Campus de Rio das Ostras
- Event
- XVII SITRAER – São Paulo, SP, Brasil, October 22–24, 2018
- Volume
- Proceedings of XVII SITRAER – Air Transportation Symposium · ISSN 2448-1912
- Pages
- 144–147
- Publisher
- Universidade de São Paulo, Escola Politécnica (POLI-USP)
- Editors
- João Batista Camargo Júnior
Abstract
In this study we consider the problem of generating long term predictions of demand for an air transport network linking a set of domestic airports so that long term capacity investments at these airports can be planned accordingly. In the considered area the concurrent as well as complementary transport modality is ground transportation. Based on demographic and economic projections, generation and attraction studies are available at different stages of the considered horizon, providing for passengers and freight flows prediction intervals. These intervals are modelized as fuzzy dual numbers allowing the development of specific modal assignment and trip distribution problems. Modal choice and traffic flows distributions are imbedded in an entropy maximization scheme which is solved along the successive stages of the planning process where traffic predictions at airports are then given by real intervals. This process can be coupled with a Dynamic Programming scheme to generate the corresponding additional capacity components to the network of airports.
Keywords: Air transport planning; networks; long term demand forecasting; entropy maximization; fuzzy intervals