Airline overbooking model based on risk analysis

Based on two kinds of risks in airline overbooking, i.e., wasting capacity and oversaling, the meaning, sources and contents of overbooking risk were discussed. Connotation of overbooking risk, risk control measures, and characteristics of revenue and cost function were highlighted. A static model and a two-step optimization dynamic model were proposed to control upper and lower risks. The static model ignored the dynamic cancellation, controlling the upper and lower risks via constraints. Combining with the measures to control the upper and lower risks, the dynamic model was taken as a two-step optimization model and was similar to a transportation programming problem which considered the substitutive capacity. The joint optimal overbooking levels of the dynamic model were computed by stochastic gradient algorithm. Results indicate that these two models have expanded the existing critical point method, and are much appropriate for decision makers to understand and to choose their preferences.