Paper Info
Title
A continuous-time seat control model for single-leg flights with no-shows and optimal overbooking upper bound
Abstract
This article proposes a continuous-time model to combine seat control and overbooking policies for single-leg flights. We assume that demand is time-and-fare dependent and follows a Poisson process. No-show passengers receive refunds which depend on their fares. Overbooking penalty is a uniformly convex function of oversale. To maximize the expected revenue, airline managers seek optimal seat allocation among competing passengers. In the meantime, they need to determine an optimal aggregate overbooking upper bound, which balances the no-show refunds and oversale penalties. Our basic model shows (i) although the nested-fare structure does not hold for the face value of fares, its essence is preserved in the sense of net fares; i.e., the face value less the no-show refund; (ii) the optimal control policy is based on a set of pre-calculated time thresholds, which is easy to implement; and (iii) there exists an optimal overbooking upper bound, below which the value function strictly increases in the upper bound, and above which the value function stays constant. We further extend the basic model to consider fare-dependent no-show rates. Numerical examples are presented.
Year
DOI
Venue
2006
10.1016/j.ejor.2005.05.008
European Journal of Operational Research
Keywords
Field
DocType
Airlines,Revenue management,Stochastic processes,Minimum acceptable fare,Aggregate overbooking upper bound
Revenue,Revenue management,Mathematical optimization,Optimal control,Upper and lower bounds,Stochastic process,Bellman equation,Convex function,Operations management,Mathematics,Rate of return
Journal
Volume
Issue
ISSN
174
2
0377-2217
Citations 
PageRank 
References 
8
0.64
15
Authors
2
Name
Order
Citations
PageRank
Youyi Feng116216.40
Baichun Xiao221738.49