Airline Revenue Management: Models for Capacity Control of a Single Leg and a Network of Flights
Author: Laila Haerian
Publisher:
Published: 2007
Total Pages: 179
ISBN-13:
DOWNLOAD EBOOKDynamic programming (DP) is one of the most powerful tools for finding the optimal booking policy for capacity control of a single leg flight. However, the extension of it to a network of flights is impractical due to the exponential growth of the size of the model with number of legs in the network. In this work we develop and use an approximate DP model to find the optimal protection levels on a single leg flight and extend it to a network of flights as well. We develop a Markov chain to calculate the expected revenue that is generated under implementation of a fixed policy at each stage of the approximate DP model and for any remaining capacity and then search for the optimal policy. We use large time chunks in the proposed DP model to decrease the computational effort and show that the resulting expected revenue converges to the expected revenue that is generated under implementation of the original DP approach. Unlike many of the existing models, in our proposed method, nesting is incorporated into the optimization procedure. Furthermore, by using the proposed Markov chain model, the expected generated revenue can be calculated directly and without using simulation.