Robust Capacity Control Choice Based Behavior

The choice-based approach to capacity control represents an attractive alternative for incorporating consumer behavior into airline revenue management. Estimation errors of choice probabilities of consumer purchase are limiting factors in applying the choice-based capacity control model to industry practice. In order to systematically mitigate the sensitivity of the optimal control policy to statistical errors, a robust capacity control model based on the finite horizon Markov decision processes (MDP) is provided. The estimate errors of choice probabilities and consumer demand are modeled by associating an uncertainty set of distributions of transition probability. This paper shows that the robust control policy can be solved efficiently. A numerical example is given to illustrate the provided robust control approach can significantly improve the worst-case expected revenue. Even if the uncertainty level is not accurately predicted, the expected revenue of the robust MDP model is mostly higher than the nominal MDP model. With a weakly condition the robust revenue is proved to increase in the remaining capacity.