Modelling for Cruise Two-Dimensional Online Revenue Management System

To solve the cruise two-dimensional revenue management problem and develop such an automated system under uncertain environment, a static model which is a stochastic integer programming is firstly constructed to maximize the total expected revenue from all kinds of cruise products. Four methods can be applied to solve the above model, which are chance constrained programming, robust optimization, deterministic programming, and bid-price control. In the chance constrained programming method, the stochastic constraints are converted into deterministic equivalent forms. In the robust optimization method, the model is transformed into a goal programming formulations with a scenario-based description of problem data. In the deterministic programming method, the stochastic demand variable is directly replaced with the mean value or expected value of demand. In the bid-price control, the rules for accepting cruise products are proposed. Further, to consider time-variable demand and increase the profit, a dynamic capacity allocation model for cruise two-dimensional revenue management is put forward by applying Markov Decision Process. Then the accept/reject optimal policies for a booking request of cruise products are obtained. The conclusions are as follows: (1) the capacity in the cruise line industry is two-dimensional, that is, the number of cabins and lifeboat seats can both affect availability of cruise products; (2) the demand for all kinds of cruise products is uncertain and the uncertainty can be coped with four solution methods; (3) the characteristics of the cruise industry and time-variable demand have to be incorporated into the static and dynamic capacity allocation models to maximize the expected revenue of the cruise line.