Modeling Passenger Disutilities in Airline Revenue Management Simulation

Passenger behavior is the fundamental factor driving air transportation market reactions to the managerial decisions of airlines; therefore, it is important to understand passenger path choice process, and to develop a valid model to represent it. In this thesis, passenger disutilities are used to indicate passengers' sensitivity to alternative path options. Accordingly, passenger disutilities have a big impact on airline revenue performances, depending on an airline's revenue management methods and the path options it provides. As an attempt to understand and represent passenger disutilities with an analytical model, this thesis describes the procedure for modeling passenger disutilities based on survey answers from airline experts. Modeling passenger disutilities assumes that they are function of market distances and that they take the form of a distribution. In this thesis, we assume that the passenger disutilities fit a linear function of market index fares in the form of Gaussian distribution for every market. In order to determine appropriate parameters for the model, the survey results obtained from airline experts are used. The coefficients of three disutility functions indicate that path quality and replanning disutilities have greater influence on passenger choice than unfavorite airline disutility does. The Passenger Origin-Destination Simulator is used to test the impact of passenger disutilities on a hypothetical 42-city, hub-and-spoke network. With all disutility functions implemented, the simulation results suggest that the role of airline revenue managements become more important with passenger preference for attractive paths. Also, the relative benefits of Origin-Destination revenue management methods as supposed to Fare Class Yield Management method are higher when passenger disutilities are considered. Among the three disutility components modeled for this thesis, the replanning disutility predominantly drives market response in our hypothetical network. Thesis Advisor: Dr. Peter P. Belobaba Title: Principal Research Scientist, Department of Aeronautics and Astronautics A