Flight Demand Forecasting Model Based On Grey Topological Prediction

The definition of airline revenue management is that the airline company can get the best benefit by selling the whole flight seats at a reasonable price, with the scientific technology of prediction and optimization. But the question is how to make all these properly. The forecast of intending flight demand is needed to allocate reservation demand reasonably. So the forecast is a very important part of the revenue management system.Based on deep analysis and wide comparison on several prediction methods, the grey topological prediction is involved to construct the flight reservation model which is mainly discussed in this paper. The first section of the paper states the revenue management system and its importance in the airline company. This is followed by the analysis of flight reservation demand with the help of its characteristic. Next, some kinds of traditional prediction methods, along with an intelligent method(BP neural network), are presented with the analysis of their shortcomings and the reasons through the simulation and comparison.Finally, the theory of grey prediction and its topological model are detailed to establish the grey topological prediction model with the reservation feature, and the predicting result can be modified to get a better purpose by the use of inter-zone prediction.The model mentioned above can be built even with much less provided data. And the data volatility can be avoided effectively, if the reservation regulation can be discovered from the change of the flight collection points. The simulation and analysis of several data groups demonstrate that the grey topological prediction has the effectiveness and practical value in the reservation demand system.