Research On Airline Seat Inventory Control Based On The Behavior Of Sell-up

Airline seat inventory control is to allocate each class rationally based on flight forecasting demands in order to maximize profits. Until now, EMSR(Expected Margin Seat Revenue) is still the basis of airline seat inventory control, but the result of EMSR depends heavily on the prediction algorithm, which assumes the independence of demand. However,more and more high-class passengers buy tickets of lower in the actual booking process,leading to differences between the types of passengers weakened. The method of revenue management based on the rate of sell up has become a future trend in order to produce the behavior of sell up. The estimation of sell up is the key of the algorithm designed, the existing algorithm is only based on historical reservation data, leading to overestimating the demands for high class. Therefore, this paper presents a new estimation algorithm of the rate of sell up based on passenger choice behavior. In the algorithm, the probability of passenger choice is introduced to model's parameters, more in line with current sales' situation.The main work of this paper can be divided into four aspects below. The first is reviewing the history and development of revenue management, pointing out the background of this topic and its significance, and briefly introducing the research status at domestic and abroad. The second is making detailed overview to the passenger arrival model, the passenger choice model, demand forecasting and airline seat inventory control model, pointing out that traditional prediction algorithm easily leads to the revenue of spiral down,and the method of revenue management based on the rate of sell up will become the tendency of future. The third is analyzing the shortcomings of the existing model of sell up, presenting a new algorithm based on passenger choice behavior and constructing an airline seat inventory control algorithm based on the behavior of sell up combined with EMSR. Finally, the paper builds simulation platform of passenger arriving and seat choice,and simulation experiments are established to verify the Availability and correctness of the algorithm.