美国数学建模竞赛特等奖论文——2002 B O The Airline Overbooking Problem【数学建模】.pdfVIP

Bumping for Dollars: The Airline Overbooking Problem 351 Bumping for Dollars: The Airline Overbooking Problem John D. Bowman Corey R. Houmard Adam S. Dickey Wake Forest University Winston-Salem, NC Advisor: Frederick H. Chen Introduction We construct a model that expresses the expected revenue for a flight in terms of the number of reservations, the capacity of the plane, the price of a ticket, the value of a voucher, and the probability of a person showing up for the flight. When values are supplied for every variable but the first, the function can be maximized to yield an optimal booking that maximizes expected revenue. We apply the model to three situations: a single flight, two flights in a chain of flights, and multiple flights in a chain of flights. We conclude that fewer flights will increase the value of the penalty or voucher and thus decrease the optimal number of reservations. Heightened security also lowers the optimal number of reservations. An increase in passengers’ fear decreases the prob- ability that a person will show up for a flight and thus increases the optimal number of reservations. Finally, the loss of billions of dollar in revenue has no effect on the optimal value of reservations. We model the probability of a given number of people showing up as a binomial distribution. We express the average expected revenue of a flight in terms of the number of bookings made. Starting with the Single-Flight case, we derive a model and revenue function for a flight unaffected by previous flights. From this situation, we expand the model to the Two-Flight case, in which the earlier flight affects the number of people who show up for the later flight. We generalize the model even further to the number of people showing up depending on many previous flights. c The UMAP Journal 351–365. Copyright 2002 by COMAP, Inc. All rights reserved. Permission to make digi